Last update: 10.04.2001 (change log)
Please check your location line carefully. If you don't see http://www.sean.de/Solaris/ in your location bar, you might want to check with the original site for the most up to date information.
SUN managed to publish a Solaris Tunable Parameters Reference Manual, applying to Solaris 8, HW 01/01. You might want to check there for anything you miss here.
nddAppendices are separate documents. They are quoted from within the text, but you might be interested in them when downloading the current document. If you say "print" for this document, the appendices will not be printed. You have to download and print them separately.
If your system behaves erratically after applying some tweaks, please don't blame me. Remember to have a backup handy before starting to tune. Always make backup copies of the files you are changing. I tried carefully to assemble the information you are seeing here, aimed at improved system performance. As usual, there are no guarantees that what worked for me will work for you. Please don't take my recommendation at heart: They are starting points, not absolutes. Always read my reasoning, don't use them blindly.
Before you start, you ought to grab a copy of the TCP state transition diagram as specified in RFC 793 on page 23. The drawback is the missing error correction supplied by later RFCs. There is an easier way to obtain blowup printouts to staple to your office walls. Grab a copy of the PostScript file pocket guide, page 2 accompanying Stevens' TCP/IP Illustrated Volume 1 [4]. Or simply open the book at figure 18.12.
I try to assemble this page and related material for everybody interested in gaining more from her or his system. If you have an item I didn't cover, but which you deem worthwhile, please write to me. A few dozen or so regular readers of this page will thank you for it. I am only human, thus if you stumble over an error, misconception, or blatant nonsense, please have me correct it. In the past, there were quite a few mistakes.
The set of documents may look a trifle colorful, or just odd, if your browser supports cascading stylesheets. Care was taken to select the formatting tags in a way that the printed output still resembles the intentions of the author, and that the set of documents is still viewable with browser like Mosaic or Lynx. Stylesheets were used as an optical enhancement. Most notable is the different color of interior and external links. Interior links are shown in greenish colors, and will be rendered within the same frame. External links on the other hand are shown in bluish colors, and all will be shown in the same new frame. If you leave it open, a new external link will be shown within the same window. Literature references within the text are often interior links, pointing to the literature section, where the external links are located.
This page and the related work have a long history in gathering. I started out peeking wide eyed over the shoulders of two people from a search engine provider when they were installing the German server of a customer of my former employer. My only alternative resource of tuning information was the brilliant book TCP/IP Illustrated 1 [4] by Stevens. I started gathering all information about tuning I was able to get my hands upon. The cumulation of these you are experiencing on these pages.
nddSolaris allows you to tune, tweak, set and reset various parameters related to the TCP/IP stack while the system is running. Back in the SunOS 4.x days, one had to change various C files in the kernel source tree, generate a new kernel, reboot the machine and try out the changes. The Solaris feature of changing the important parameters on the fly is very convenient.
Many of the parameters I mention in the rest of the document you are reading are time intervals. All intervals are measured in milliseconds. Other parameters are usually bytecounts, but a few times different units of measurements are used and documented. A few items appear totally unrelated to TCP/IP, but due to the lack of a better framework, they materialized on this page.
Most tunings can be achieved using the program ndd. Any
user may execute this program to read the current settings, depending on
the readability of the respective device files. But only the super user is
allowed to execute ndd -set to change values. This makes sense
considering the sensitive parameters you are tuning. Details on the use of
ndd can be obtained from the respective manual page.
ndd will become your friend, as it is the major tool to
tweak most of the parameters described in this document. Therefore you
better make yourself familiar with it. A quick overview will be given in
this section, too. ndd is not limited to tweaking TCP/IP
related parameters. Many other devices, which have a device file underneath
/dev and a kernel module can be configured with the help of
ndd. For instance, any networking driver which supports the
Data Link Provider Interface (DLPI) can be configured.
The parameters supplied to ndd are symbolic keys indexing
either a single usually numerically value, or a table. Please note that the
keys usually (but not always) start out with the module or device name. For
instance, changing values of the IP driver, you have to use the device file
/dev/ip and all parameters start out with ip_.
The question mark is the most notable exception to this rule.
The interactive mode allows you to inspect and modify a device, driver or module interactively. In order to inspect the available keyword names associated with a parameter, just type the question mark. The next item will explain about the output format of the parameter list.
# ndd /dev/tcp name to get/set ? tcp_slow_start_initial value ? length ? 2 name to get/set ? ^D
The example above queries the TCP driver for the value of the slow start feature in an interactive fashion. The typed input is shown boldface.
If you are interested in the parameters you can tweak for a given
module, query for the question mark. This special parameter name is part
of all ndd configurable material. It tells the names of all
parameters available - including itself - and the access mode of the
parameter.
# ndd /dev/icmp \? ? (read only) icmp_wroff_extra (read and write) icmp_def_ttl (read and write) icmp_bsd_compat (read and write) icmp_xmit_hiwat (read and write) icmp_xmit_lowat (read and write) icmp_recv_hiwat (read and write) icmp_max_buf (read and write)
Please mind that you have to escape the question mark with a backslash from the shell, if you are querying in the non-interactive fashion as shown above.
At the command line, you often need to check on settings of your TCP/IP stack or other parameters. By supplying the parameter name, you can examine the current setting. It is permissible to mention several parameters to check on at once.
# ndd /dev/udp udp_smallest_anon_port 32768 # ndd /dev/hme link_status link_speed link_mode 1 1 1
The first example checks on the smallest anonymous port UDP may use when sending a PDU. Please refer to the appropriate section later in this document on the recommended settings for this parameter.
The second example checks the three important link report values of a 100 Mbit ethernet interface. The results are separated by an empty line, because some parameters may refer to tabular values instead of a single number.
This mode of interaction with ndd will frequently be
found in scripts or when changing value at the command line in a
non-interactive fashion. Please note that you may only set one value
at a time. The scripts section below contains
examples in how to make changes permanent using a startup script.
# ndd -set /dev/ip ip_forwarding 0
The example will stop the forwarding of IP PDUs, even if more than one non-local interface is active and up. Of course, you can only change parameters which are marked for both, reading and writing.
Andres Kroonmaa kindly supplied a nifty script to check all existing values for a network component (tcp, udp, ip, icmp, etc.). Usually I do the same thing using a small Perl script.
This document is separated into several chapters with little inter-relation. It is still advisable to loosely follow the order outlined in the table of contents.
The first chapter entirely focusses on the TCP connection queues. It is quite long for such small topic, but it is also meant to introduce you into my style of writing. The next chapter deals with TCP retransmission related parameters that you can adjust to your needs. The chapter is more concise. One chapter on deals with path MTU discovery, as there used to be problems with older versions of Solaris. Recent versions usually do not need any adjustments.
The fifth chapter is a kind of catch-all. Some TCP, some UDP and some IP related parameters are explained (forwarding, port ranges, timers), and a quick detour into bug 1226653 explains that some versions were capable of sending packages larger than the MTU. The following chapter in depth deals with windows, buffers and related issues.
Chapter seven detours from the ndd interface, and focusses
on variables you can set in your /etc/system file, as some
things can only be thus managed. Another part of that chapter deals with
the hme interface and appropriate tunables. The chapter may be
split in future, and parts of it are already found in the appendices.
The chapter dealing with patches, an important topic with any OS, just points you to various sources, and only mentions some essential things for older versions of Solaris.
Literature exists in abundance. The literature sections is more a lose collection of links and some books that I consider essential when working with TCP/IP, not limited to Solaris. The RFC sections is kind of hard to keep up-to-date, but then, I reckon you know how to read the rfc-index file.
The final chapters quickly glance at new or at one time new versions of
Solaris - time makes them obsolete. The chapter is there for historical
reason, more or less. The scripts sections deals with the
nettune script used by YaSSP. It finishes with some TODO material.
This section is dedicated exclusively to the various queues and tunable
variable(s) used during connection instantiation. The socket API maintains
some control over the queues. But in order to tune anything, you have to
understand how listen and accept interact with
the queues. For details, see the various Stevens books mentioned in the
literature section.
listen, the kernel moves the socket from
the TCP state CLOSED into the state LISTEN, thus
doing a passive open. All TCP servers work like this. Also, the kernel
creates and initializes various data structures, among them the socket buffers and two queues:
This queue contains an entry for every SYN that has
arrived. BSD sources assign so_q0len entries to this
queue. The server sends off the ACK of the client's
SYN and the server side SYN. The connection
get queued and the kernel now awaits the completion of the TCP three
way handshake to open a connection. The socket is in the
SYN_RCVD state. On the reception of the client's
ACK to the server's SYN, the connection stays
one round trip time (RTT) in this queue before the kernel
moves the entry into the
This queue contains an entry for each connection for
which the three way handshake is completed. The socket is in the
ESTABLISHED state. Each call to accept()
removes the front entry of the queue. If there are no entries in the
queue, the call to accept usually blocks. BSD source
assign a length of so_qlen to this queue.
Both queues are limited regarding their number of entries. By calling
listen(), the server is allowed to specify the size of the
second queue for completed connections. If the server is for whatever
reason unable to remove entries from the completed connection
queue, the kernel is not supposed to queue any more connections. A
timeout is associated with each received and queued SYN
segment. If the server never receives an acknowledgment for a queued
SYN segment, TCP state SYN_RCVD, the time will
run out and the connection thrown away. The timeout is an important
resistance against SYN flood attacks.
 |
 |
|
| Figure 1: Queues maintained for listening sockets. | Figure 2: TCP three way handshake, connection initiation. |
Historically, the argument to the listen function specified the maximum number of entries for the sum of both queues. Many BSD derived implementations multiply the argument with a fudge factor of 3/2. Solaris <= 2.5.1 do not use the fudge factor, but adds 1, while Solaris 2.6 does use the fudge factor, though with a slightly different rounding mechanism than the one BSD uses. With a backlog argument of 14, Solaris 2.5.1 servers can queue 15 connections. Solaris 2.6 server can queue 22 connections.
Stevens shows that the incomplete connection queue does need
more entries for busy servers than the completed
connection queue. The only reason for specifying a large backlog value is
to enable the incomplete connection queue to grow as SYN
arrive from clients. Stevens shows that moderately busy webserver has an
empty completed connection queue during 99 % of the time, but the
incomplete connection queue needed 15 or less entries in 98 % of
the time! Just try to imagine what this would mean for a really busy
webcache like Squid.
Data for an established connection which arrives before the connection is
accept()ed, should be stored into the socket buffer. If the
queues are full when a SYN arrived, it is dropped in the hope
that the client will resend it, hopefully finding room in the queues
then.
According to Cockroft [2], there was only one listen queue for unpatched Solari <= 2.5.1. Solari >= 2.6 or an applied TCP patch 103582-12 or above splits the single queue in the two shown in figure 1. The system administrator is allowed to tweak and tune the various maxima of the queue or queues with Solaris. Depending on whether there are one or two queues, there are different sets of tweakable parameters.
The old semantics contained just one tunable parameter
tcp_conn_req_max which specified the maximum argument for
the listen(). The patched versions and Solaris 2.6 replaced
this parameter with the two new parameters
tcp_conn_req_max_q0 and
tcp_conn_req_max_q. A SunWorld article on 2.6 by Adrian
Cockroft tells the following about the new parameters:
tcp_conn_req_max [is] replaced. This value is well-known as it normally needs to be increased for Web servers in older releases of Solaris 2. It no longer exists in Solaris 2.6, and patch 103582-12 adds this feature to Solaris 2.5.1. The change is part of a fix that prevents denial of service from
SYNflood attacks. There are now two separate queues of partially complete connections instead of one.tcp_conn_req_max_q0 is the maximum number of connections with handshake incomplete. A
SYNflood attack could only affect this queue, and a special algorithm makes sure that valid connections can still get through.tcp_conn_req_max_q is the maximum number of completed connections waiting to return from an accept call as soon as the right process gets some CPU time.
In other words, the first specifies the size of the incomplete connection queue while the second parameters assigns the maximum length of the completed connection queue. All three parameters are covered below.
You can determine if you need to tweak this set of parameters by
watching the output of netstat -sP tcp. Look for the value of
tcpListenDrop, if available on your version of Solaris. Older
versions don't have this counter. Any value showing up might indicate
something wrong with your server, but then, killing a busy server (like
squid) shuts down its listening socket, and might increase this counter
(and others). If you get many drops, you might need to increase the
appropriate parameter. Since connections can also be dropped, because
listen() specifies a too small argument, you have to be
careful interpreting the counter value. On old versions, a SYN
flood attack might also increase this counter.
Newer or patched versions of Solaris, with both queues available, will
also have the additional counters tcpListenDropQ0 and
tcpHalfOpenDrop. Now the original counter
tcpListenDrop counts only connections dropped from the
completed connection queue, and the counter ending in
Q0 the drops from the incomplete connection queue.
Killing a busy server application might increase either or both counters.
If the tcpHalfOpenDrop shows up values, your server was likely
to be the victim of a SYN flood. The counter is only
incremented for dropping noxious connection attempts. I have no idea, if
those will also show up in the Q0 counter, too.
The current parameter describes the maximum number of pending connection requests queued for a listening endpoint in the completed connection queue. The queue can only save the specified finite number of requests. If a queue overflows, nothing is sent back. The client will time out and (hopefully) retransmit.
The size of the completed connection queue does not influence
the maximum number of simultaneous established connections after they were
accepted nor does it have any influence on
the maximum number of clients a server can serve. With Solaris, the maximum
number of file descriptors is the limiting factor for simultaneous
connections, which just happened to coincide with the maximum backlog queue
size.
From the viewpoint of TCP those connections placed in the completed
connection queue are in the TCP state ESTABLISHED, even
though the application has not reaped the connection with a call to
accept. That is the number limited by the size of the queue,
which you tune with this parameter. If the application, for some reason,
does not release entries from the queue by calling accept, the
queue might overflow, and the connection is dropped. The client's TCP will
hopefully retransmit, and might find a place in the queue.
Solaris offers the possibility to place connections into the backlog
queue as soon as the first SYN arrives, called eager
listening. The three way handshake will be completed as soon as the
application accept()s the connection. The use of eager
listening is not recommended for production systems.
Solari < 2.5 have a maximum queue length of 32 pending connections. The length of the completed connection queue can also be used to decrease the load on an overloaded server: If the queue is completely filled, remote clients will be denied further connections. Sometimes this will lead to a connection timed out error message.
Naively, I assumed that a very huge length might lead to a long service time on a loaded server. Stevens showed that the incomplete connection queue needs much more attention than the completed connection queue. But with tcp_conn_req_max you have no option to tweak that particular length.
Earlier versions of this document suggested to tune tcp_conn_req_max with regards to the values of rlim_fd_max and rlim_fd_cur, but the interdependencies are more complex than any rule of thumb. You have to find your own ideal. When a connection is still in the queue, only the queue length limits the number of entries. Connections taken from the queue are put into a file descriptor each.
There is a trick to overcome the hardcoded limit of 1024 with a patch. SunSolve shows this trick in connection with
SYNflood attacks. A greatly increased listen backlog queue may offer some small increased protection against this vulnerability. On this topic also look at the tcp_ip_abort_cinterval parameter. Better, use the mentioned TCP patches, and increase the q0 length.echo "tcp_param_arr+14/W 0t10240" | adb -kw /dev/ksyms /dev/memThis patch is only effective on the currently active kernel, limiting its extend to the next boot. Usually you want to append the line above on the startup script
/etc/init.d/inetinit. The shown patch increases hard limit of the listen backlog queue to 10240. Only after applying this patch you may use values above 1024 for the tcp_conn_req_max parameter.
A further warning: Changes to the value of
tcp_conn_req_max parameter in a running system
will not take effect until each listening application is
restarted. The backlog queue length is evaluated whenever an application
calls listen(3N), usually once during startup. Sending a HUP
signal may or may not work; personally I prefer to TERM the application and
restart them manually or, even better, use a startup script.
After installing the mentioned TCP patches, alternatively after installing Solaris 2.6, the parameter tcp_conn_req_max is no longer available. In its stead the new parameters tcp_conn_req_max_q and tcp_conn_req_max_q0 emerged. tcp_conn_req_max_q0 is the maximum number of connections with handshake incomplete, basically the length of the incomplete connection queue.
In other words, the connections in this queue are just being
instantiated. A SYN was just received from the client, thus
the connection is in the TCP SYN_RCVD state. The connection
cannot be accept()ed until the handshake is complete, even if
the eager listening is active.
To protect against SYN flooding, you can increase this parameter. Also refer to the parameter tcp_conn_req_max_q above. I believe that changes won't take effect unless the applications are restarted.
After installing the mentioned TCP patches, alternatively after installing Solaris 2.6, the parameter tcp_conn_req_max is no longer available. In its stead the new parameters tcp_conn_req_max_q and tcp_conn_req_max_q0 emerged. tcp_conn_req_max_q is the length of the completed connection queue.
In other words, connections in this queue of length
tcp_conn_req_max_q have completed the three way handshake
of a TCP open. The connection is in the state ESTABLISHED.
Connections in this queue have not been accept()ed by the
server process (yet).
Also refer to the parameter tcp_conn_req_max_q0. Remember that changes won't take effect unless the applications are restarted.
This parameter specifies the minimum number of available connections
in the completed connection queue for select()
or poll() to return "readable" for a listening (server)
socket descriptor.
Programmers should note that Stevens [7] describes a
timing problem, if the connection is RST between the
select() or poll() call and the subsequent
accept() call. If the listening socket is blocking, the
default for sockets, it will block in accept() until a
valid connection is received. While this seems no tragedy with a
webserver or cache receiving several connection requests per second,
the application is not free to do other things in the meantime, which
might constitute a problem.
The retransmission timeout values used by Solaris are way too aggressive for wide area networks, although they can be considered appropriate for local area networks. SUN thus did not follow the suggestions mentioned in RFC 1122. Newer releases of the Solaris kernel are correcting the values in question:
The recommended upper and lower bounds on the RTO are known to be inadequate on large internets. The lower bound SHOULD be measured in fractions of a second (to accommodate high speed LANs) and the upper bound should be 2*MSL, i.e., 240 seconds.
Besides the retransmit timeout (RTO) value two further parameters R1 and R2 may be of interest. These don't seem to be tunable via any Solaris' offered interface that I know of.
The value of R1 SHOULD correspond to at least 3 retransmissions, at the current RTO. The value of R2 SHOULD correspond to at least 100 seconds.
[...]
However, the values of R1 and R2 may be different for SYN and data segments. In particular, R2 for a SYN segment MUST be set large enough to provide retransmission of the segment for at least 3 minutes. The application can close the connection (i.e., give up on the open attempt) sooner, of course.
Great many internet servers which are running Solaris do retransmit segments unnecessarily often. The current condition of European networks indicate that a connection to the US may take up to 2 seconds. All parameters mentioned in the first part of this section relate to each other!
As a starter take this little example. Consider a picture, size 1440
byte, LZW compressed, which is to be transferred over a serial linkup with
14400 bps and using a MTU of 1500. In the ideal case only one PDU gets
transmitted. The ACK segment can only be sent after
the complete PDU is received. The transmission takes about 1 second. These
values seem low, but they are meant as 'food for thought'. Now consider
something going awry...
Solaris 2.5.1 is behaving strange, if the initial SYN
segment from the host doing the active open is lost. The initial
SYN gets retransmitted only after a period of 4 *
tcp_rexmit_interval_initial plus a constant C. The time is 12
seconds with the default settings. More information is being prepared on
the retransmission test page.
The initial lost SYN may or may not be of importance in
your environment. For instance, if you are connected via ATM SVCs, the
initial PDU might initiate a logical connection (ATM works point to point)
in less than 0.3 seconds, but will still be lost in the process. It is
rather annoying for a user of 2.5.1 to wait 12 seconds until something
happens.
This interval is waited before the last data sent is retransmitted due to a missing acknowledgment. Mind that this interval is used only for the first retransmission. The more international your server is, the larger you should chose this interval.
Special laboratory environments working in LAN-only environments might be better off with 500 ms or even less. If you are doing measurements involving TCP (which is almost always a bad idea), you should consider lowering this parameter.
Why do I consider TCP measurements a bad idea? If ad-hoc approaches are used, or there is no deeper knowledge of the mechanics of TCP, you are bound to arrive at wrong conclusions. Unless there are TCP dumps to document that indeed what you expect is actually happening, results may lead to wrong conclusions. If done properly, there is nothing wrong with TCP measurements. The same rules apply, if you are measuring protocols on top of TCP.
There are lots of knobs and dials to be fiddled with - all of which need to be documented along with the results. Scientific experiments need to be repeatable by others in order to verify your findings.
After the initial retransmission further retransmissions will start after the tcp_rexmit_interval_min interval. BSD usually specifies 1500 milliseconds. This interval should be tuned to the value of tcp_rexmit_interval_initial, e.g. some value between 50 % up to 200 %. The parameter has no effect on retransmissions during an active open, see my accompanying document on retransmissions.
The tcp_rexmit_interval_min doesn't display any influence on connection establishment with Solaris 2.5.1. It does with 2.6, though. The influence on regular data retransmissions, or FIN retransmissions I have yet to research.
This interval specifies how long retransmissions for a connection
in the ESTABLISHED state should be tried before a
RESET segment is sent. BSD systems default to 9 minutes.
This interval specifies how long retransmissions for a remote host are
repeated until the RESET segment is sent. The difference to
the tcp_ip_abort_interval parameter is that this
connection is about to be established - it has not yet reached the state
ESTABLISHED. This value is interesting considering
SYN flood attacks on your server. Proxy server are doubly
handicapped because of their Janus behavior (like a server towards the
downstream cache, like a client towards the upstream server).
According to Stevens this interval is connected to the active
open, e.g. the connect(3N) call. But according to
SunSolve the interval has an impetus on both
directions. A remote client can refuse to acknowledge an opening connection
up to this interval. After the interval a RESET is sent. The
other way around works out, too. If the three-way handshake to open a
connection is not finished within this interval, the RESET
Segment will be sent. This can only happen, if the final ACK
went astray, which is a difficult test case to simulate.
To improve your SYN flood resistance, SUN suggests to use
an interval as small as 10000 milliseconds. This value has only been tested
for the "fast" networks of SUN. The more international your connection is,
the slower it will be, and the more time you should grant in this interval.
Proxy server should never lower this value (and should let Squid
terminate the connection). Webservers are usually not affected, as they
seldom actively open connections beyond the LAN.
All previously mentioned retransmissions related interval use an exponential backoff algorithm. The wait interval between two consecutive retransmissions for the same PDU is doubled starting with the minimum.
The tcp_rexmit_interval_max interval specifies the maximum wait interval between two retransmissions. If changing this value, you should also give the abort interval an inspection. The maximum wait interval should only be reached shortly before the abort interval timer expires. Additionally, you should coordinate your interval with the value of tcp_close_wait_interval or tcp_time_wait_interval.
This parameter specifies the timeout before sending a delayed
ACK. The value should not be increased above
500, as required by
RFC 1122. This value is of great
interest for interactive services. A small number will increase the
"responsiveness" of a remote service (telnet, X11), while a larger value
can decrease the number of segments exchanged.
The parameter might also interest to HTTP servers which transmit small
amounts of data after a very short retrieval time. With a heavy-duty
servers or in laboratory banging environment, you might encounter service
times answering a request which are well above 50 ms. An increase to 500
might lead to less PDUs transferred over the network, because TCP is able
to merge the ACK with data. Increases beyond 500 should
not be even considered.
SUN claims that Solaris recognizes
the initial data phase of a connection. An initial ACK (not
SYN) is not delayed. As opposed to the simplistic
approach mentioned in the SUN paper, a request for a webservice (both,
server or proxy) which does not fit into a single PDU can be transmitted
faster. Also check the
tcp_slow_start_initial Parameter.
The tcp_deferred_ack_interval also seems to be used to distinguish full-sized segments between interactive traffic and bulk data transfer. If a sender uses MSS sized segments, but sends each segment further apart than approximately 0.9 times the interval, the traffic will be rated interactive, and thus every segment seems to get ACKed.
This parameter features the maximum number of segments received after
which an ACK just has to be sent. Previously I thought this parameter
solely related to interactive data transfer, but I was mistaken. This
parameter specifies the number of outstanding ACKs. You can
give it a look when tuning for high speed traffic and bulk transfer, but
the parameter is controversial. For instance, unless you employ selective
acknowledgments (SACK) like Solaris 7, you can only ACK the number of
segments correctly received. With the parameter at a larger value,
statistically the amount of data to retransmit is larger.
Good values for retransmission tuning don't beam into existence from a white source. Rather you should carefully plan an experiment to get decent values. Intervals from another site can not be carried over to another Solaris system without change. But they might give you an idea where to start when choosing your own values.
The next part looks at a few parameters having to do with retransmissions, as well.
This parameter provides the slow-start bug discovered in BSD and Windows
TCP/IP implementations for Solaris. More information on the topic can be
found on the servers of
SUN and in Stevens [6]. To summarize the effect, a server starts
sending two PDUs at once without waiting for an ACK due to
wrong ACK counts. The ACK from connection
initiation being counted as data ACK - compare with figure 2. Network congestion avoidance algorithms are
being undermined. The slow start algorithm does not allow the buggy
behavior, compare with
RFC 2001.
Setting the parameter to 2 allows a Solaris machine to behave like it has the slow start bug, too. Well, IETF is said to make amends to the slow start algorithm, and the bug is now actively turned into a feature. SUN also warns:
It's still conceivable, although rare, that on a configuration that supports many clients on very slow-links, the change might induce more network congestions. Therefore the change of tcp_slow_start_initial should be made with caution.
[...]
Future Solaris releases are likely to default to 2.
You can also gain performance, if many of your clients are running old BSD or derived TCP/IP stacks (like MS). I expect new BSD OS releases not to figure this bug, but then I am not familiar with the BSD OS family. A reader of this page told me about cutting the latency of his server in half, just by using the value of 2.
If you want to know more about this feature and its behavior, you can have a look at some experiments I have conducted concerning that particular feature. The summary is that I agree with the reader: A BSDish client like Windows definitely profits from using a value of 2.
I reckon that this parameter deals with the slow start for an already established connection which was idle for some time (however the term idle is defined here).
Something to do with the number of duplicates ACKs. If we
do fast retransmit and fast recovery algorithms, this many
ACKs must be retransmitted until we assume that a segment has
really been lost. A simple reordering of segments usually causes no more
than two duplicate ACKs.
There are a couple of parameters which require some elementary familiarity with RFC 2001, which covers TCP Slow Start, Congestion Avoidance, Fast Retransmit, and Fast Recovery Algorithms, as well as ssthresh and cwnd.
This parameter controls when things like rtt_sa (the smoothed RTT), rtt_sd (the smoothed mean deviation), and ssthresh (the slow start threshold) are cached in the routing table. By default, Solaris does not cache any of the parameters. It is claimed that you can set it to a value you like, but to be the same as BSD, use 16.
The value to this parameter is the number of RTT samples that had to be sampled, so that an accurate enough value can be stored in the routing table. If you chose to use this feature, use a value of 16 or above. Using 16 allows the smoothed RTT filter to converge within 5 % of the correct value, compare Stevens [4], chapter 21.9.
The parameters may do more than described here. If a routing table entry is not directly connected and not being used, the cache for things like rtt_sa, rtt_sd and ssthresh associated with the entry will be flushed after 30 seconds. The parameter tcp_rtt_updates must be greater than zero to enable the cache.
I could imagine that external helper programs invoked by MRTG on a regular basis connecting to a far-away host might benefit from increasing this value slightly above the invocation interval.
Whenever a connection is about to be established, the three-way handshake open negotiation, the segment size used will be set to the minimum of (a) the smallest MTU of an outgoing interface, and (b) from MSS announced by the peer. If the remote peer does not announce a MSS, usually the value 536 will be assumed. If path MTU discovery is active, all outgoing PDUs have the IP option DF (don't fragment) set.
If the ICMP error message fragmentation needed is received, a router on the way to the destination needed to fragment the PDU, but was not allowed to do so. Therefore the router discarded the PDU and did send back the ICMP error. Newer router implementations enclose the needed MSS in the error message. If the needed MSS is not included, the correct MSS must be determined by trial and error algorithm.
Due to the internet being a packet switching network, the route a PDU travels along a TCP virtual circuit may change with time. For this reason RFC 1191 recommends to rediscover the path MTU of an active connection after 10 minutes. Improvements of the route can only be noticed by repeated rediscoveries. Unfortunately, Solaris aggressively tries to rediscover the path MTU every 30 seconds. While this is o.k. for LAN environments, it is a grossly impolite behavior in WANs. Since routes may not change that often, aggressive repetitions of path MTU discoveries leads to unnecessary consumption of channel capacity and elongated service times.
Path MTU discovery is a far reaching and controversial topic when discussing it with local ISPs. Still, pMTU discovery is at the foundation of IPv6. The PSC tuning page argues pro path MTU discovery, especially if you maintain a high-speed or long-delay (e.g. satellite) link.
The recommendation I can give you is not to use the defaults of Solaris < 2.5. Please use path MTU discovery, but tune your system RFC conformant. You may alternatively want to switch off the path MTU discovery all together, though there are few situations where this is necessary.
I was made aware of the fact that in certain circumstances bridges connecting data link layers of differing MTU sizes defeat pMTU discovery. I have to put some more investigation into this matter. If a frame with maximum MTU size is to be transported into the network with the smaller MTU size, it is truncated silently. A bridge does not know anything about the upper protocol levels: A bridge neither fragments IP nor sends an ICMP error.
There may be work-arounds, and the tcp_mss_def is one of them. Setting all interfaces to the minimum shared MTU might help, at the cost of losing performance on the larger MTU network. Using what RFC 1122 calls an IP gateway is a possible, yet expensive solution.
This timer determines the interval Solaris rediscovers the path MTU. An extremely large value will only evaluate the path MTU once at connection establishment.
This parameter switches path MTU discovery on or off. If you enter a 0 here, Solaris will never try to set the DF bit in the IP option - unless your application explicitly requests it.
This is a debug switch! When activated, this switch will have the IP or TCP layer ignore all ICMP error messages fragmentation needed. By this, you will achieve the opposite of what you intended.
This parameter determines the default MSS (maximum segment size) for non-local destination. For path MTU discovery to work effectively, this value can be set to the MTU of the most-used outgoing interface descreased by 20 byte IP header and 20 byte TCP header - if and only if the value is bigger than 536.
Solaris 8 supports IPv6. Since IPv6 uses different defaults for the maximum segment size, one has to distinguish between IPv4 and IPv6. The default for IPv6 is close to what is said for tcp_mss_def.
This section covers a variety of topics, starting with various TCP timers which do not relate to previously mentioned issues. The next subsection throws a quick glance at some erratic behavior. The final section looks at a variety of parameters which deal with the reservation of resources.
Additionally, I strongly suggest the use of a file /etc/init.d/nettune (always called
first script) which changes the tunable parameters.
/etc/rcS.d/S31nettune is a hardlink to this file. The script
will be executed during bootup when the system is in single user
mode. A killscript is not necessary. The section about startup scripts below reiterates this topic in greater
depth.
The current subsection covers three important TCP timers. First I will
have a look at the keepalive timer. The timer is rather controversial, and
some Solari implement them incorrectly. The next parameter limits the
twice maximum segment lifetime (2MSL) value, which is connected to
the time a socket spends in the TCP state TIME_WAIT. The final
entry looks at the time spend in the TCP state FIN_WAIT_2.
This value is one of the most controversial ones when talking with other people about appropriate values. The interval specified with this key must expire before a keep-alive probe can be sent. Keep-alive probes are described in the host requirements RFC 1122: If a host chooses to implement keep-alive probes, it must enable the application to switch them on or off for a connection, and keep-alive probes must be switched off by default.
Keep-alives can terminate a perfectly good connection (as far
as TCP/IP is concerned), cost your money and use up transmission capacity
(commonly called bandwidth, which is, actually, something completely
different). Determining whether a peer is alive should be a task of the
application and thus kept on the application layer. Only if you run into
the danger of keeping a server in the ESTABLISHED state
forever, and thus using up precious server resources, you should switch on
keep-alive probes.
Figure 3: A typical handshake during a transaction.
Figure 3 shows the typical handshake during a HTTP connection. It is of no importance for the argumentation if the server is threaded, preforked or just plain forked. Webservers work transaction oriented as is shown in the following simplified description - the numbers do not relate to the figure:
Common implementations need to exchange 9..10 TCP segments per HTTP connection. The keep-alive option as a HTTP/1.0 protocol and extensions can be regarded as a hack. Persistent connections are a different matter, and not shown here. Most people still use HTTP/1.0, especially the Squid users.
The keep-alive timer becomes significant for webservers, if in step 1 the client crashed or terminates without the server knowing about it. This condition can be forced sometimes by quickly pressing the stop button of netscape or the Logo of Mosaic. Thus the keep-alive probes do make sense for webservers. HTTP Proxies look like a server to the browser, but look like a client to the server they are querying. Due to their server like interface, the conditions for webservers are true for proxies, as well.
With an implementation of keep-alive probes working
correctly, a very small value can make sense when trying to
improve webservers. In this case you have to make sure that the probes stop
after a finite time, if a peer does not answer. Solari <= 2.5
have a bug and send keep-alive probes forever. They seem to want
to elicit some response, like a RST or some ICMP error message
from an intermediate router, but never counted on the destination simply
being down. Is this fixed with 2.5.1? Is there a patch available against
this misbehavior? I don't know, maybe you can help me.
I am quite sure that this bug is fixed in 2.6 and that it is safe to use a small value like ten minutes. Squid users should synchronize their cache configuration accordingly. There are some Squid timeouts dealing with an idle connection.
Even though the parameter key contains "close_wait" in its name, the value specifies the TIME_WAIT interval! In order to fix this kind of confusion, starting with Solaris 7, the parameter tcp_close_wait_interval was renamed to the correct name tcp_time_wait_interval. The old key tcp_close_wait_interval still exists for backward compatibility reasons. User of Solari below 7 must use the old name tcp_close_wait_interval. Still, refer to tcp_time_wait_interval for an in-depth explaination.
As Stevens repeatedly states in his books, the TIME_WAIT
state is your friend. You should not desperately try to avoid it, rather
try to understand it. The maximum segment lifetime(MSL) is the
maximum interval a TCP segment may live in the net. Thus waiting twice this
interval ensures that there are no leftover segments coming to haunt you.
This is what the 2MSL is about. Afterwards it is safe to reuse the socket
resource.
The parameter specifies the 2MSL according to the four minute limit specified in RFC 1122. With the knowledge about current network topologies and the strategies to reserve ephemerical ports you should consider a shorter interval. The shorter the interval, the faster precious resources like ephemerical ports are available again.
A toplevel search engine implementor recommends a value of 1000 millisecond to its customers. Personally I believe this is too low for regular server. A loaded search engine is a different matter alltogether, but now you see where some people start tweaking their systems. I rather tend to use a multiple of the tcp_rexmit_interval_initial interval. The current value of tcp_rexmit_interval_max should also be considered in this case - even though retransmissions are unconnected to the 2MSL time. A good starting point might be the double RTT to a very remote system (e.g. Australia for European sites). Alternatively a German commercial provider of my acquaintance uses 30000, the smallest interval recommended by BSD.
This values seems to describe the (BSD) timer interval which prohibits a
connection to stay in the FIN_WAIT_2 state forever.
FIN_WAIT_2 is reached, if a connection closes actively. The
FIN is acknowledged, but the FIN from the passive
side didn't arrive yet - and maybe never will.
Usually webservers and proxies actively close connections - as long as you don't use persistent connection and even those are closed from time to time. Apart from that HTTP/1.0 compliant server and proxies close connections after each transaction. A crashed or misbehaving browser may cause a server to use up a precious resource for a long time.
You should consider decreasing this interval, if netstat -f
inet shows many connections in the state FIN_WAIT_2.
The timer is only used, if the connection is really idle. Mind
that after a TCP half close a simplex data transmission is still available
towards the actively closing end. TCP half closes are not yet supported by
Squid, though many web servers do support them (certain HTTP drafts suggest
an independent use of TCP connections). Nevertheless, as long as the client
sends data after the server actively half closed an established connection
the timer is not active.
Sometimes, a Squid running on Solaris (2.5.1) confuses the system
utterly. A great number of connection to a varying degree are in
CLOSE_WAIT for reasons beyond me. During this phase the proxy
is virtually unreachable for HTTP requests though, obnoxiously, it still
answers ICP requests. Although lowering the value for tcp_close_wait_interval
is only fixing symptoms indirectly, not the cause, it may help overcoming
those periods of erratic behavior faster than the default. The thing needed
would be some means to influence the CLOSE_WAIT interval
directly.
I noticed that Solari < 2.6 behave erratically under some conditions,
if the IPX ethernet MTU of 1500 is used. Maybe there is an error in the
frame assembly algorithm. If you limit yourself to the IEEE 802.3 MTU of
1492 byte, the problem does not seem to appear. A sample startup script with link in /etc/rc2.d can
be used to change the MTU of ethernet interfaces after their
initialization. Remember to set the MTU for every virtual interface,
too!
Note, with a patched Solaris 2.5.1 or Solaris 2.6, the problem does not seem to appear. Limiting your MTU to non-standard might introduce problems with truncated PDUs in certain (admittedly very special) environments. Thus you may want to refrain from using the above mentioned script (always called second script in this document).
Since I observed the erratic behavior only in a Solaris 2.5, I believe it has been fixed with patch 103169-10, or above. The error description reads "1226653 IP can send packets larger than MTU size to the driver."
The following parameters have little impact on performance, nevertheless I reckon them worth noting here. Please note that parameters starting with the ip6 prefix apply to IPv6 while its twin with the ip applies to IPv4:
If you intend to disable the routing abilities of your host all together, because you know you don't need them, you can set this switch to 0. The default value of 2 was only available in older versions of Solaris. It activates IP forwarding, if two or more real interfaces are up. The value of 1 in Solari < 8 activates IP forwarding regardless of the number of interfaces. With the possible exception of MBone routers and firewalling, you should leave routing to the dedicated routing hardware.
Starting with Solaris 8, the parameter set is split. You use ip_forwarding and ip6_forwarding to overall switch on forwarding of IPv4 and IPv6 PDU respectively between interfaces. The interfaces participating in forwarding can be activated separately, see if:ip_forwarding. Unless you host is acting as router, it is still recommended for security reasons to switch off any forwarding between interfaces.
Please replace the if part of the parameter name with the
appropriate interface available on your system, e.g. hme0 or
hme0. Look into the available /dev/ip parameters,
if unsure what interfaces are known to the IP stack.
Starting with Solaris 8, a subset of interfaces participating in IP forwarding can be selected by setting the appropriate parameter to 1. You also need to set the ip6_forwarding and ip_forwarding parameter, if you want to forward IPv6 or IPv6 respectively.
For security reasons, and in many environments, forwarding is not recommended.
This parameter determines if IP datagrams can be forwarded which have the source routing option activated. The parameter has little meaning for performance but is rather of security relevance. Solaris may forward such datagrams, if the host route option is activated, bypassing certain security construct - possibly undermining your firewall. Thus you should disable it always, unless the host functions as a regular router (and no other services).
If you enabled IPv6 forwarding or IPv4 forwarding, the *_forward_src_routed parameters may relate to forwarding.
This switch decides whether datagrams directed to any of your direct broadcast addresses can be forwarded as link-layer broadcasts. If the switch is on (default), such datagrams are forwarded. If set to zero, pings or other broadcasts to the broadcast address(es) of your installed interface(s) are silently discarded. The switch is recommended for any host, but can break "expected" behavior.
If you don't want to respond to an ICMP echo request (usually generated
by the ping program) to any of your IPv4 broadcast or IPv6
multicast addresses addresses, set the matching parameter to 0. On one
hand, responding to broadcast pings is rumored to have caused panics, or at
least partial network meltdowns. On the other hand, it is a valid behavior,
and often used to determine the number of alive hosts on a particular
network. If you are dead sure that neither you nor your network admin will
need this feature, you can switch it off by using the value of 0.
If you do not want to respond to any IPv4-broadcast or IPv6-multicast probes for security reasons, it is recommended to set the matching parameter to 0.
Solaris IP only generates ip_icmp_err_burst ICMP error messages in any ip_icmp_err_interval, regardless of IPv4 or IPv6. In order to protect from denial of service (DOS) attacks, the parameters do not need to be changed. Some administrators may need a higher error generation rate, and thus may want to decrease the interval or increase the generated message.
In versions of Solaris prior to 8, ip_icmp_err_interval used to define the minimum time between two consecutive ICMP error responses - as if in older versions the (by then not existing) ip_icmp_err_burst parameter had a value of 1. The generated ICMP responses include the time exceeded message as evoked by the traceroute command. If your current setting here is above the RTT of a traceroute probe, usually the second probe you see will time out.
If you set ip_icmp_err_burst to exactly 0, traceroute will not give away your host as running Solaris. Also, you switched of the rate limitation of ICMP messages, and are thus open to DOS attacks. Of course, there are other ways to determine which TCP/IP implementation a networked host is running.
The parameters control the number of bytes returned by any ICMP error message generated on this Solaris host. The default value 64 is sufficient for most cases. Some laboratory environments may want to temporarily increase the value in order to figure out problems with some network services.
These parameters control whether the IPv4 or IPv6 part of the IP stack send ICMP redirect messages. For security reasons, it is recommended to disable sending out such messages, unless your host is acting as router.
If you enabled IPv6 forwarding or IPv4 forwarding, the *_send_redirects parameters may relate to forwarding.
This flag control, if your routing table can be updated by ICMP redirect messages. Unless you run your host to act as router, it is recommended to disable this feature for security reasons. Otherwise, malicious external hosts may confuse your routing table.
If you enabled IPv6 forwarding or IPv4 forwarding, the *_ignore_redirects parameters may relate to forwarding.
This parameter limits the number of virtual interfaces you can declare per physical interface. Especially if you run Web Polygraph, you will need to increase the number of virtual interfaces available on your system.
According to RFC 1122, a host is said to be multihomed, if it has more than one IP address. Each IP address is assumed to be a logical interface. Different logical interfaces may map to the same physical interface. Physical interfaces may be connected to the same or different networks.
The strong end system model aka strict multihoming requires a host not to accept datagrams on physical interfaces to which to logical one is not bound. Outgoing datagrams are restricted to the interface which corresponds with the source ip address.
The weak end system model aka loose multihoming lets a host accept any of its ip addresses on any of its interfaces. Outgoing datagrams may be sent on any interface.
For security reasons, it is recommended to require strict multihoming, that is, setting the parameter to value 1. In certain circumstances, though, it may be necessary to disable strict multihoming, e.g. if the host is connected to a virtual private networks (VPN) or sometimes when acting as firewall.
For instance, I once maintained a setup, where a pair of related caching proxies were talking exclusively to each other via a crossover cable on one interface using private addresses while the other interface was connected to the public internet. In order to have them actually use the behind-the-scenes link, I had to manually set routes and disable strict multihoming.
There are some parameters related to the ranges of ports associated with reserved access and non-privileged access. This section deals with the majority of useful parameters when selecting different than default port ranges.
This value has the same size for UDP and TCP. Solaris allocates ephemerical ports above 32768. Busy servers or hosts using a large 2MSL, see tcp_close_wait_interval, may want to lower this limit to 8192. This yields more precious resources, especially for proxy servers.
A contra-indication may be servers and services running on well known ports above 8192. This parameter should be set very early during system bootup, especially before the portmapper is started.
The IANA port numbers document requires the assigned and/or private ports to start at 49152. For busy servers, severly limiting their ephemerical port supply in such a manner is not an option.
This parameter has to be seen in combination with
udp_smallest_anon_port. The traceroute
program tries to reach a random UDP port above 32768 - or rather tries not
to reach such a port - in order to provoke an ICMP error message from the
host.
Paranoid system administrator may want to lower the value for this reason down to 32767, after the corresponding value for udp_smallest_anon_port has been lowered. On the other hand, datagram application protocols should be able to cope with foreign protocol datagrams.
If an ICP caching proxy or other UDP hyper-active applications are used, the lowering of this value can not be recommended. The respective TCP parameter tcp_largest_anon_port does not suffer this problem.
The largest anonymous port for TCP should be the largest possible port number. There is no need to change this parameter.
Privileged ports can only be bound to by the superuser. The smallest non-privileged port is the first port that a regular user can have his or her application to bind to.
The extra priviledged ports are those priviledged ports outside the scope of the reserved ports. Reserved port numbers are usually below 1024, see tcp_smallest_nonpriv_port for TCP and tcp_smallest_nonpriv_port for UDP, and require superuser privileges in order to bind to. For instance, if NFS is activated, the NFS server port 2049 is marked as privileged.
You can examine the extra privileged TCP port by looking at the read-only parameter tcp_extra_priv_ports. If you need to add an extra privileged port, use the tcp_extra_priv_ports_add with the port number as argument. If you need to remove an extra privileged port, use the tcp_extra_priv_ports_del action with the port number to remove as parameter. You can only add or remove one port at a time.
# ndd /dev/tcp tcp_extra_priv_ports 2049 4045 # ndd -set /dev/tcp tcp_extra_priv_ports_add 4444 5555 # ndd /dev/tcp tcp_extra_priv_ports 2049 4045 4444 # ndd -set /dev/tcp tcp_extra_priv_ports_del 4444 # ndd /dev/tcp tcp_extra_priv_ports 2049 4045
Analogous procedures apply to UDP extra privileged port.
This section is about windows, buffers and watermarks. It is still work in progress. The explanations available to me were very confusing (sigh), though the Stevens [7] helped to clear up a few things. If you have corrections to this section, please let me know and contribute to an update of the page. Many readers will thank you!
Figure 4: buffers and related issues
Here just a short trip through the network layer in order to explain what happens where. Your application is able to send almost any size of data to the transport layer. The transport layer is either UDP or TCP. The socket buffers are implemented on the transport layer. Depending on your choice of transport protocol, different actions are taken on this level.
Only when the data was acknowledged from the peer instance, the data can be removed from the socket buffer! For slow connections or a slowly working peer, this implies a very long time some data uses up the buffer.
Please assume that there is not really a socket buffer for sending UDP. This really depends on the operating systems, but many systems copy the user data to some kernel storage area, whereas others try to eliminate all copy operations for the sake of performance.
Please note that for the reverse direction, that is receiving datagrams, UDP does indeed employ real buffering.
The IP layer needs to fragment chunks which are too large. Among the reasons TCP prechunks its segments is the need to avoid fragmentation. IP searches the routing tables for the appropriate interface in order to determine the fragment size and interface.
If the output queue of the datalink layer interface is full, the datagram will be discarded and an error will be returned to IP and back to the transport layer. If the transport protocol was TCP, TCP will try to resend the segment at a later time. UDP should return the ENOBUFS error, but some implementations don't.
To determine the MTU sizes, use the ifconfig -a command.
The MTUs are needed for some calculation to be done later in this section.
With IPv4 you can determine the MSS from the interface MTU by subtracting
20 Bytes for the TCP header and 20 Bytes for the IP header. Keep this in
mind, as the calculation will be repeatedly necessary in the text following
below.
$ ifconfig -a
lo0: flags=849<UP,LOOPBACK,RUNNING,MULTICAST> mtu 8232
inet 127.0.0.1 netmask ff000000
hme0: flags=863<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST> mtu 1500
inet 130.75.3.xxx netmask ffffff80 broadcast 130.75.3.255
ci0: flags=843<UP,BROADCAST,RUNNING,MULTICAST> mtu 9180
inet 130.75.214.xxx netmask ffffff00 broadcast 130.75.214.255
ether xx:xx:xx:xx:xx:xx
fa0: flags=842<BROADCAST,RUNNING,MULTICAST> mtu 9188
inet 0.0.0.0 netmask 0
ether xx:xx:xx:xx:xx:xx
el0: flags=843<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 130.75.215.xxx netmask ffffff00 broadcast 130.75.215.255
ether xx:xx:xx:xx:xx:xx
I removed the uninteresting things. hme0 is the regular 100 Mbps ethernet interface. The 10 Mbps ethernet interface is called le0. The el0 interface is an ATM LAN emulation (lane) interface. ci0 is the ATM classical IP (clip) interface. fa0 is the interface that supports Fore's proprietary implementation of native ATM. Fore is the vendor of the installed ATM card. AFAIK you can use this interface to build PVCs or, if you are also using Fore switches, SVCs. You see an unconfigured interface there.
The buffer sizes for sending and receiving TCP segment and for UDP
datagrams can be tuned with Solaris. With the help of the
netstat command you can obtain an output similar but unlike
the following one. The data was obtained on a server which runs a Squid
with five dnsserver children. Since the interprocess communication is
accomplished via localhost sockets, you see both, the client side and the
server side of each dnsserver child socket.
$ netstat -f inet TCP Local Address Remote Address Swind Send-Q Rwind Recv-Q State -------------------- -------------------- ----- ------ ----- ------ ------- blau-clip.ssh challenger-clip.1023 57344 19 63980 0 ESTABLISHED localhost.38437 localhost.38436 57344 0 57344 0 ESTABLISHED localhost.38436 localhost.38437 57344 0 57344 0 ESTABLISHED localhost.38439 localhost.38438 57344 0 57344 0 ESTABLISHED localhost.38438 localhost.38439 57344 0 57344 0 ESTABLISHED localhost.38441 localhost.38440 57344 0 57344 0 ESTABLISHED localhost.38440 localhost.38441 57344 0 57344 0 ESTABLISHED localhost.38443 localhost.38442 57344 0 57344 0 ESTABLISHED localhost.38442 localhost.38443 57344 0 57344 0 ESTABLISHED localhost.38445 localhost.38444 57344 0 57344 0 ESTABLISHED localhost.38444 localhost.38445 57344 0 57344 0 ESTABLISHED
The columns titled with Swind and Rwind
contain values for the size of the respective send- and reception
windows, based on the free space available in the receive
buffer at each peer. The Swind column
contains the offered window size as reported by the remote
peer. The Rwind column displays the advertised
window size being transmitted to the remote peer.
An application can change the size of the the socket layer
buffers with calls to setsockopt with the
parameter SO_SNDBUF or SO_RCVBUF. Windows and
buffers are not interchangeable. Just remember: The buffers have a fixed
size - unless you use setsockopt to change. Windows on the
other hand depend on the free space available in the input buffer. The
minimum and maximum requirements for buffer sizes are tunable
watermarks.
Figure 5: buffers, watermarks and window sizes.
Figure 5 shows the relation of the different buffers, windows and watermarks. I decided to let the send buffer grow from the maximum towards zero, which is just a way of showing things, and does probably not represent the real implementation. I left out the different socket options as the picture is confusing enough.
Effectively the size of the receive buffer. The receive high watermark
tunes the transport layer socket buffer sizes on a kernel wide basis. The
socket option SO_RCVBUF allows the dynamic change of the
receive buffer size within the application on a per socket basis.
The number of maximum sized segments which have to fit into a receive buffer. This is a k.o.-criterion for sizing buffers. It is not possible to set a receive buffer size which is unable to have at least this many segments fit into it.
minimum amount of data in the input buffer to have a call to
select or poll return the socket as readable.
The socket option
SO_RCVLOWAT allows the dynamic change of the receive low
watermark on a per socket basis. With UDP, the socket is reported readable
as soon as there is a complete datagram in the receive buffer.
Effectively the size of the send buffer. The send buffer high watermark
tunes the transport layer socket buffer size on a kernel wide basis. The
socket buffer can also be changed on a per-socket basis by using the
SO_SNDBUF socket option within an application. Mind that for
UDP the size of the output buffer represents the maximum datagram size.
The amount of free space to be available in the send buffer to have
select and poll report the socket writable. The
socket option SO_SNDLOWAT allows a dynamic change of this size
on a per-socket basis.
The offered window is the amount of free space available in the input
buffer of the peer (how much we are allowed to send), relative to the
acknowledged data. The window size is shown in the Swind
column in the netstat output. From the offered window, the
usable window is calculated, that is the amount of data which can be send
as soon as possible. TCP never sends more than the minimum of the current
congestion window and the offered window.
to_send := MIN( cwnd, offered window )
The advertised window is the free space available in the local input
buffer advertised to the remote peer (how much we are willing to
receive). The advertised window is shown in the Rwind column
in the netstat output.
The congestion window cwnd is used in TCP bulk data transfer with the congestion avoidance and slow start algorithms. The cwnd is initialized to 1 segment, and will grow with received acknowledgments. It will not exceed the maximum congestion window size. TCP will not send more than the minimum of the offered window and the current value of cwnd.
Maximum size the congestion window may grow to. Usually, the maximum size is larger than the buffer size available, which is not a problem.
Squid users should note the following behavior seen with Solaris 2.6. The default socket buffer sizes which are detected during configuration phase are representative of the values for tcp_recv_hiwat, udp_recv_hiwat, tcp_xmit_hiwat and udp_xmit_hiwat. Also note that enabling the hit object feature still limits hit object size to 16384 byte, regardless of what your system is able to achieve.
Output from Squid 1.1.19 configuration script on a Solaris 2.6 host with the previously mentioned parameters all set to 64000. Please mind that these parameters do not constitute optimal sizes in most environments:
checking Default UDP send buffer size... 64000 checking Default UDP receive buffer size... 64000 checking Default TCP send buffer size... 64000 checking Default TCP receive buffer size... 64000
Buffers and windows are very important if you link via satellite. Due to the daterate possible but the extreme high round-trip delays of a satellite link, you will need very large TCP windows and possibly the TCP timestamp option. Only RFC 1323 conformant systems will achieve these ends. In other words, get a Solaris 2.6. For 2.5 systems, RFC 1323 compliance can be purchased as a Sun Consulting Special.
Window sizes are important for maximum throughput calculations, too. As Stevens [4] shows, you cannot go faster than the window size offered by your peer, divided by the round-trip time (RTT). The lower your RTT, the faster you can transmit. The larger your window, the faster you can transmit. If you intend to employ maximum window sizes, you might want to give tcp_deferred_acks_max another look.
The network research laboratory of the German research network did measurements on satellite links. The RTT for a 10 Mbps link (if I remember correctly) was about 500 ms. A regular system was able to transmit 600 kbps whereas a RFC 1323 conformant system was able to transmit about 7 Mbps. Only bulk data transfer will do that for you.
(1) 10 Mbps * 0.5 s = 5 Mbit = 625 KB (2) 512 KB / 0.5 s = 1 MBps = 8 Mbps (3) 64 KB / 0.5 s = 128 KBps = 1 Mbps
The bandwidth-delay-product can be used to estimate the initial value when tweaking buffer sizes. The buffers then represent the capacity of the link. If we apply the bandwidth-delay-product calculations to the satellite link above, we get the following results: Equation 1 estimates the buffer sizes necessary to fully fill the 10 Mbps link. Equation 2 assumes that the buffer sizes were set to 512 KB, which would yield 8 Mbps. Slight deviation in the experiment may have been caused by retransmissions. Finally, equation 3 estimates the maximum datarate we can use on the satellite link, if limited to 64 KB buffers, e.g. Solaris <= 2.5.1. The 1 Mbps constitute an upper limit, as can be seen by the measured 600 Kbps.
Application developers, especially those for web-based applications, should be aware of the implications of persistent connections. As long as HTTP/1.0 connection-per-transaction style is used by your application, depending on the size of the transaction data, you will not get any decent transmissions via satellite. For instance, the average web object is about 13 KByte in size, thus transmitting such an object on a connection-per-transaction basis will never get past TCP slow start. While this may or may not be a big deal with terrestrial links, but you will never be able to fill a satellite pipe to a satisfactorily degree. Doing things in parallel might help. Only when reaching TCP congestion avoidance you will see any filling of the pipe. You might also want to check out the unrelated tcp_slow_start_initial parameter.
A word of caution seems to be in order, when tuning the Solaris' TCP high watermarks: Starting with Solaris 2.6, setting tcp_xmit_hiwat or tcp_recv_hiwat near 65535 may have the side effect of turning on the wscale option, because these values are rounded up to multiples of MTU for each connection. In some cases you may not want to accidentally use wscale, because it may break something else in your setup such as IP-Filter. To avoid accidentally using wscale, you need to make sure that tcp_xmit_hiwat and tcp_recv_hiwat are both at least 1 MTU below 65535. For ethernet interfaces, 64000 is a good choice.
This parameter describes the maximum size the congestion window can be opened. The congestion window is opened as large as possible with any Solaris up to 2.5.1. A change to this value is only necessary for older Solaris systems, which defaulted to 32768. The Solaris 2.6 default looks reasonable, but you might need to increase this further for satellite or long, fast links.
Though window sizes beyond 64k are possible, mind that the window scale option is only announced during connection creation and your maximum windows size is 1 GByte (1,073,725,440 Byte). Also, the window scale option is only employed during the connection, if both sides support it.
This parameter determines the maximum size of the initial TCP reception buffer. The specified value will be rounded up to the next multiple of the MSS. From the free space within the buffer the advertised window size is determined. That is, the size of the reception window advertised to the remote peer. Squid users will be interested in this value with regards to the socket buffer size the Squid auto configuration program finds.
The previous table shows an Rwind value of 63980 = 7 *
9140. 9140 is the MSS of the ATM classical IP interface (clip) in host
blau. The interface itself uses a MTU of 9180. For the standard
builtin 10 Mbps or 100 Mbps IPX ethernet, you get a MTU of 1500 on the
outgoing interface, which yields an MSS of 1460. The value of 57344 in the
next Rwind line points to the lo0 (loopback)
interface, MTU 8232, MSS 8192 and 57344 = 7 * 8192.
Starting with Solaris 2.6 values above 65535 are possible, see the window scale option from RFC 1323. Only if the peer host also implements RFC 1323, you will benefit from buffer sizes above 65535. If one host does not implement the window scale option, the window is still limited to 64K. The option is only activated, if buffer sizes above 64K are used.
For HTTP, I don't see the need to increase the buffer above 64k. Imagine servicing 1024 simultaneous connections. If both the TCP high watermarks of your system are tuned to 64k and your application uses the system's defaults, you would need 128M just for your TCP buffers!
Squid's configuration option tcp_recv_bufsize lets you
select a TCP receive buffer size, but if set to 0 (default) the kernel
value will be taken, which is configurable with the
tcp_recv_hiwat parameter. A buffer size of 16K is large
enough to cover over 70 % of all received webobjects on our caches.
Refer to tcp_host_param for a way to configure special defaults for a set of hosts and networks.
This parameter influences the minimum size of the input buffer. The reception buffer is at least as large as this value multiplied by the MSS. The real value is the maximum of tcp_recv_hiwat round up to the next MSS and tcp_recv_hiwat_minmss multiplied by the MSS, in other words, something akin to:
hiwat_tmp ~= ceil( tcp_recv_hiwat / MSS ) real_size := MAX( hiwat_tmp, tcp_recv_hiwat_minmss ) * MSS
That way, however bad you misconfigure the buffers, there is a guaranteed space for tcp_recv_hiwat_minmss full segments in your input buffer.
The highwater mark for the UDP reception buffer size. This value may be of interest for Squid proxies which use ICP extensively. Please read the explanations for tcp_recv_hiwat. Squid users will want at least 16384, especially if you are planning on using the (obsolete) hit object feature of Squid. A larger value lets your computer receive more seemingly simultaneous ICP PDUs.
If you see many dead parent detections in your cache.log
file without cause, you might want to increase the receive buffer. In most
environments an increase to 64000 will have a negligible effect on the
memory consumption, as most application, including Squid, use only one or
very few UDP sockets, and often in an iterative way.
Remember if you don't set your socket buffer explicitly with a call to
setsockopt(), your default reception buffer will have about
the mentioned size. Arriving Datagrams of a larger size might be truncated
or completely rejected. Some systems don't even notify your receiving
application.
This parameter influence a heuristic which determines the size of the initial send window. The actual value will be rounded up to the next multiple of the MSS, e.g. 8760 = 6 * 1460. Also do read the section on tcp_recv_hiwat.
The table further to the top shows a Swind of 57344 = 7 *
8192. For the standard builtin 10 Mbps or 100 Mbps IPX ethernet, you get an
MTU of 1500 on the outgoing interface, which yields a MSS of 1460.
Starting with Solaris 2.6 values above 65535 are possible, see the window scale option from RFC 1323. Only if the peer host also implements RFC 1323, you will benefit from buffer sizes above 65535. If one host does not implement the window scale option, the window is still limited to 64K.
I don't see the need to increase the buffer above 32K for HTTP applications. Imagine servicing 1024 simultaneous connections. If both TCP high watermarks of your system are tuned to 32K, you would need 64M just for your TCP buffers. Mind that the send buffer has to keep a copy of all unacknowledged segments. Therefore it is affordable to give it a greater size than the receive buffer. Again, 16K covers over 70 % of all transferred web objects on our caches, and 32K should cover 90 %.
Refer to tcp_host_param for a way to configure special defaults for a set of hosts and networks.
This refers to the highwater mark for send buffers. May be of interest for proxies using ICP extensively. Please refer to the explanations for tcp_xmit_hiwat. Squid users will want 16384, especially if you are planning on using the hit object feature of Squid. Selecting a higher value for the transmission is not feasible.
Please remember that there exists no real send buffer for UDP on the
socket layer. Thus, trying to send a larger amount of data than
udp_xmit_hiwat will truncate the excess, unless the
SO_SNDBUF socket option was used to extend the allowed
size.
The current parameter refers to the amount of data which must be
available in the TCP socket sendbuffer until select or
poll return writable for the connected file
descriptor.
Usually there is no need to tune this parameter. Applications can use
the socket option SO_SNDLOWAT to change this parameter on a
process local basis.
The current parameter refers to the amount of data which must be
available until select or poll return
writable for the connected file descriptor. Since UDP does not
need to keep datagrams and thus needs no outgoing socket buffer, the socket
will always be writable as long as the socket sendbuffer size value is
greater than the low watermark. Thus it does not really make much sense to
wait for a datagram socket to become writable unless you constantly adjust
the sendbuffer size.
Usually there is no need to tune this parameter, especially not on a system-wide basis.
Finally found the explanations in the SUN TCP/IP Admin Guide. The current parameter refers to the maximum buffer size an application is allowed to specify with the SO_SNDBUF and SO_RCVBUF socket option calls. Attempts to use larger buffers will fail with a EINVAL return code from the socket option call. SUN recommends to use only the largest buffer necessary for any of your applications - that is, the supremum function, not the sum. Specifying a greater size does not seem to have much impact, if all your applications are well-behaving. If not, they may consume quite an amount of kernel memory, thus this parameter is also a kind of safety line.
A few odd remarks at this point, concerning the recommendations given for the transmission buffer sizes. I decreased the recommendations of Adrian Cockroft in favor of a more conservative memory consumption. Also, with an average HTTP object size of 13 KByte, you can expect to fit over 50 % of all objects into the transmission buffer. On the other hand, larger objects which are to be transmitted by a cache or webserver may suffer in certain circumstances. Furthermore, I should recommend a generic transmission buffer size which is double the reception buffer size. This recommendation bases on the fact that unacknowledged segments occupy the send buffer until they are acknowledged.
Here some more material from the SUN TCP/IP Admin Guide, kindly pointed out by Mr. Murphy. Refer to the SUN guide for a more detailed description of these parameters, and their respective applicability. Most noteworthy is tcp_host_param, which allows per host/network defaults regarding RFC 1323 TCP options.
If the parameter is set (non-zero), then the TCP window scale option will always be negotiated during connection initiation. Otherwise, the scale option will only be used if the buffer size is above 64K. To take effect, both hosts have to support RFC 1323.
If the parameter is set (non-zero), then the TCP timestamp option will always be negotiated during connection initiation. The scale option will always be used if the remote system sent a timestamp option during connection initiation. To use the timestamp, both hosts have to support RFC 1323.
If the option is set (non-zero), the TCP timestamp option will be used in addition to the TCP window scale option, if the user has requested a buffer size above 64K, that is, if window scaling is active.
Refer to tcp_host_param for instructions on handling the table. The same rules apply except that the ipv6 table is meant for IPv6, of course.
This parameter represents a table which contains special TCP options to
be used with a remote host or network. The table is configurable with the
help of ndd, and empty by default. The following piece of code
displays the contents of the table at various points, sets an entry and
removes it again:
# ndd /dev/tcp tcp_host_param Hash HSP Address Subnet Mask Send Receive TStamp # ndd -set /dev/tcp tcp_host_param '192.168.4.17 sendspace 262144 recvspace 262144' # ndd /dev/tcp tcp_host_param Hash HSP Address Subnet Mask Send Receive TStamp 125 62bae844 192.168.004.017 000.000.000.000 0000262144 0000262144 0 # ndd -set /dev/tcp tcp_host_param '192.168.4.17 delete' # ndd /dev/tcp tcp_host_param Hash HSP Address Subnet Mask Send Receive TStamp
Use the mask command to supply a netmask for a network, and the timestamp command to supply the timestamp option. Fill this table from a startup script, if you want large default windows only for certain links (e.g. which go via satellite), but small windows for anything else. The content of this table takes precedence over the generic global values, if certain criteria are met:
This section evolved around tuning items, which were not directly related to the TCP/IP stack, but nevertheless play an important role in the tuning of any system. Refer to SUN's Solaris Tunable Reference Manual for more in-depth information.
Did you reserve enough swap space? You should have at least
as much swap as you have main memory. If you have little main
memory, even double your swap. Do not be fooled by the result
of the vmstat command - read the manpage and realize that the
small value for free memory shown there is (usually) correct.
With Solaris there seems to exist a difference between virtually
generated processes and real processes. The latter is extremely dependent
on the amount of virtual memory. To test the amount of both kinds of
processes, try a small program of mine. Do start it
at the console, without X and not as privileged user. The first
value is the hard limit of processes, and the second value the amount of
processes you can really create given your virtual memory configuration.
Tweaking your ulimit values may or may not help.
/etc/systemThe file /etc/system contains various very important
resource configurable parameters for your system. You use these tunings to
give a heavily loaded system more resources of a certain kind.
Unfortunately a reboot is necessary after changing anything.
Though one could schedule reboots after midnight, I advice against it. You
should always check if your changes have the desired effect, and won't tear
down the system.
Adrian Cockroft severely warns against transporting an
/etc/system from one system onto another, even worse, onto
another hardware platform:
/etc/system when you upgrade.
The most frequent changes are limited to the number of file descriptors, because the socket API uses file descriptors for handling internet connectivity. You may want to look at the hard limit of filehandles available to you. Proxies like Squid have to count twice to thrice for each request: open request descriptors and an open file and/or (depending what squid you are using) an open forwarding request descriptors. Similar calculations are true for other caches.
You are able to influence the tuning with the reserved word
set. Use a whitespace to separate the key from the keyword.
Use an equals sign to separate the value from its key. There are a few
examples in the comments of the file.
Please, before you start, make a backup copy of your initial
/etc/system. The backup should be located on your
root filesystem. Thus, if some parameters fail, you can always supply the
alternative, original system file on the boot prompt. The following shows
two typically entered parameters:
* these are the defaults of Solaris < 8 set rlim_fd_max=1024 set rlim_fd_cur=64
WARNING! SUN does not make any guarantees for the correct working of your system, if you use more file descriptors than 4096. Personally, my old fvwm window manager did quit working alltogether. In my case, I compiled it on a Solaris 2.3 or 2.4 system and transferred it always onwards to a 2.5 system. After re-compiling it on the new OS, it worked to my satisfaction.
If you experience SEGV core dumps from your select(3c)
system call after increasing your file descriptors above 4096, you have to
recompile the affected programs. Especially the select(3c)
call is known to the Squid users for its bad tempers concerning the maximum
number of file descriptors. SUN remarks to this topic:
Note: This does not work as expected! See text below.
I did test this suggestion by SUN, and a friend of mine tried it with
Squid Caches. The result was a complete success or disaster both times,
depending on your point of view: If you can live with supplying naked women
to your customers instead of bouncing logos of companies, go ahead and try
it. If you really need to access file descriptors above 1024, don't
use select(), use poll() instead!
poll() is supposed to be faster with Solaris, anyway. A
different source mentions that the redefinition workaround mentioned above
works satisfactorily; not for me, my personal experiences warn against such
an action.
At the pages of VJ are a some tricks which I incorporated into this paper, too. Personally I am of the opinion that the VJ pages are not as up to date as they could be.
Many parameters of interest can be determined using the sysdef
-i command. Please keep in mind that many values are in
hexadecimal notation without the 0x prefix. Another very
good program to see your system's configuration is sysinfo, the program. Refer to the manpages how to
invoke this program.
There is also the possibility to use a small helper script kindly supplied by Mr. Kroonma to have a look into some kernel variables with the help of the absolute debugger (adb). You can extend the script to suit your own needs, but you should know what you are doing. Refer to the manual page of the absolute debugger for details of displaying non-ulong datatype variables. If you don't know, what adb can do for you, hands off.
This parameters defines the soft limit of open files you can have. The currently active soft limit can be determined from a shell with something like
ulimit -Sn
Use at your own risk values above 256, especially if you are running old binaries. A value of 4096 may look harmless enough, but may still break old binaries.
Another source mentions that using more than 8192 file descriptors is discouragable. It mentions that you ought to use more processes, if you need more than 4096 file descriptors. On the other hand, an ISP of my acquaintance is using 16384 descriptors to his satisfaction.
The predicate rlim_fd_cur <= rlim_fd_max must be fulfilled.
Please note that Squid only cares about the hard limit (next item). With respect to the standard IO library, you should not raise the soft limit above 256. Stdio can only use <= 256 FDs. You can either use AT&T'ssfio library, or use Solaris 64-bit mode applications which fix the stdio weakness. RPC prior to 2.6 may break, if more than 1024 FDs are available to it.
Also note that RPC prior to Solaris 2.6 may break, if more than 1024 FDs are available to it. Also, setting the soft limit to or above 1024 implies that your license server queries break (first hand experience - thanks Jens). Using 256 is really a strong recommendation.
This parameter defines the hard limit of open files you can have. For a Squid and most other servers, regardless of TCP or UDP, the number of open file descriptors per user process is among the most important parameter. The number of file descriptors is one limit on the number of connections you can have in parallel. You can find out the value of your hard limit on a shell with something like
ulimit -Hn
You should consider a value of at least 2 * tcp_conn_req_max and you should provide at least 2 * rlim_fd_cur. The predicate rlim_fd_cur <= rlim_fd_max must be fulfilled.
Use at your own risk values above 1024. SUN does not make any
warranty for the workability of your system, if you increase this above
1024. Squid users of busy proxies will have to increase this value, though.
A good starting seems to be 16384 <= x <= 32768. Remember to change
the Makefile for Squid to use poll() instead of
select(). Also remember that each call of
configure will change the Makefile back, if you didn't change
Makefile.in.
Any decent application will incorporate code to increase its soft limit to a possibly higher hard limit. Please note (again) that Squid, as such an application, only cares about the hard limit.
sd driver with wide-SCSI), 1048576 (SPARC storage
array driver), no recommendations
A work-copy of this value is often stored in the mount structure or driver structure at the time it is attached. If a driver sees IO requests larger than this parameter, the requests will be broken down into appropriotely sized chunks. The file system may further fragment the chunks. A change might be conceivable, if your database server uses raw devices and issues large requests - mind that many of todays database usage paradigms result in many small chunked requests and will not speed up by increasing this value.
If working large chunked IO with UFS, you can additionally increase the number of cylinder groups and decrease the number of inodes per group (as there will be a few large files).
This parameter determines the size of certain kernel data structures which are initialized at startup. Recent versions of Solaris derive most table sizes now from the amount of memory available, but there are still some dependent variables on this parameter, see max_nprocs, maxuprc, ufs_ninode, ncsize and ndquot. There is strong indication that the default for maxusers itself is being determined from the main memory in megs. It might also be a function of the available memory and/or architecture.
The greater you chose the number for maxusers, the greater the number of the mentioned resources. The relation in strictly proportional: A doubling of maxusers will (more or less) double the other resources.
Adrian Cockroft advises against a setting of maxusers. The kernel uses a lot of space while keeping track of the RAM usages within the system, therefore it might need to be reduced on systems with gigabytes of main memory. The point to change this parameter is whenever the automagically determined number of user processes is way too high, e.g. file servers, database servers, compute servers with few processes, or way too low.
Starting with Solaris 8, you can determine the number of the largest
possible value for a pid_t the system can set. From this
parameter, the kernel variable maxpid will be set
once during startup. maxpid on the other hand cannot
be set via /etc/system.
This parameter is the mysterious difference between the number of all
processes max_nprocs and the number of user processes
maxuprc, and affects the number of system process table
slots reserved for uid 0, e.g. sched, pageout
and fsflush.
Though a change is not immanently recommended, increasing the number of root slots to 10 plus number of root processes might be considered, in order to provide root with a shell at times the system is uncapable of creating a user-level shell, e.g. run-away user-processes, fork-of-death, etc.
This is the systemwide number of processes available, user and system processes. You should leave sufficient space to the parameter maxuprc. The value of this parameter is influenced by the setting of maxusers.
The number is used to compute various further parameters (see below), including the DNLC cache, the quota structures, System-V semaphore limits, address translation table resources for sun4m, sun4d and Intel Solaris verions.