1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
|
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
<HTML>
<HEAD>
<TITLE>Connections in FIN_WAIT_2 and Apache</TITLE>
<LINK REV="made" HREF="mailto:marc@apache.org">
</HEAD>
<!-- Background white, links blue (unvisited), navy (visited), red (active) -->
<BODY
BGCOLOR="#FFFFFF"
TEXT="#000000"
LINK="#0000FF"
VLINK="#000080"
ALINK="#FF0000"
>
<!--#include virtual="header.html" -->
<H1 ALIGN="CENTER">Connections in the FIN_WAIT_2 state and Apache</H1>
<OL>
<LI><H2>What is the FIN_WAIT_2 state?</H2>
Starting with the Apache 1.2 betas, people are reporting many more
connections in the FIN_WAIT_2 state (as reported by
<CODE>netstat</CODE>) than they saw using older versions. When the
server closes a TCP connection, it sends a packet with the FIN bit
sent to the client, which then responds with a packet with the ACK bit
set. The client then sends a packet with the FIN bit set to the
server, which responds with an ACK and the connection is closed. The
state that the connection is in during the period between when the
server gets the ACK from the client and the server gets the FIN from
the client is known as FIN_WAIT_2. See the <A
HREF="ftp://ds.internic.net/rfc/rfc793.txt">TCP RFC</A> for the
technical details of the state transitions.<P>
The FIN_WAIT_2 state is somewhat unusual in that there is no timeout
defined in the standard for it. This means that on many operating
systems, a connection in the FIN_WAIT_2 state will stay around until
the system is rebooted. If the system does not have a timeout and
too many FIN_WAIT_2 connections build up, it can fill up the space
allocated for storing information about the connections and crash
the kernel. The connections in FIN_WAIT_2 do not tie up an httpd
process.<P>
<LI><H2>But why does it happen?</H2>
There are numerous reasons for it happening, some of them may not
yet be fully clear. What is known follows.<P>
<H3>Buggy clients and persistent connections</H3>
Several clients have a bug which pops up when dealing with
<A HREF="../keepalive.html">persistent connections</A> (aka keepalives).
When the connection is idle and the server closes the connection
(based on the <A HREF="../mod/core.html#keepalivetimeout">
KeepAliveTimeout</A>), the client is programmed so that the client does
not send back a FIN and ACK to the server. This means that the
connection stays in the FIN_WAIT_2 state until one of the following
happens:<P>
<UL>
<LI>The client opens a new connection to the same or a different
site, which causes it to fully close the older connection on
that socket.
<LI>The user exits the client, which on some (most?) clients
causes the OS to fully shutdown the connection.
<LI>The FIN_WAIT_2 times out, on servers that have a timeout
for this state.
</UL><P>
If you are lucky, this means that the buggy client will fully close the
connection and release the resources on your server. However, there
are some cases where the socket is never fully closed, such as a dialup
client disconnecting from their provider before closing the client.
In addition, a client might sit idle for days without making another
connection, and thus may hold its end of the socket open for days
even though it has no further use for it.
<STRONG>This is a bug in the browser or in its operating system's
TCP implementation.</STRONG> <P>
The clients on which this problem has been verified to exist:<P>
<UL>
<LI>Mozilla/3.01 (X11; I; FreeBSD 2.1.5-RELEASE i386)
<LI>Mozilla/2.02 (X11; I; FreeBSD 2.1.5-RELEASE i386)
<LI>Mozilla/3.01Gold (X11; I; SunOS 5.5 sun4m)
<LI>MSIE 3.01 on the Macintosh
<LI>MSIE 3.01 on Windows 95
</UL><P>
This does not appear to be a problem on:
<UL>
<LI>Mozilla/3.01 (Win95; I)
</UL>
<P>
It is expected that many other clients have the same problem. What a
client <STRONG>should do</STRONG> is periodically check its open
socket(s) to see if they have been closed by the server, and close their
side of the connection if the server has closed. This check need only
occur once every few seconds, and may even be detected by a OS signal
on some systems (<EM>e.g.</EM>, Win95 and NT clients have this capability, but
they seem to be ignoring it).<P>
Apache <STRONG>cannot</STRONG> avoid these FIN_WAIT_2 states unless it
disables persistent connections for the buggy clients, just
like we recommend doing for Navigator 2.x clients due to other bugs.
However, non-persistent connections increase the total number of
connections needed per client and slow retrieval of an image-laden
web page. Since non-persistent connections have their own resource
consumptions and a short waiting period after each closure, a busy server
may need persistence in order to best serve its clients.<P>
As far as we know, the client-caused FIN_WAIT_2 problem is present for
all servers that support persistent connections, including Apache 1.1.x
and 1.2.<P>
<H3>A necessary bit of code introduced in 1.2</H3>
While the above bug is a problem, it is not the whole problem.
Some users have observed no FIN_WAIT_2 problems with Apache 1.1.x,
but with 1.2b enough connections build up in the FIN_WAIT_2 state to
crash their server.
The most likely source for additional FIN_WAIT_2 states
is a function called <CODE>lingering_close()</CODE> which was added
between 1.1 and 1.2. This function is necessary for the proper
handling of persistent connections and any request which includes
content in the message body (<EM>e.g.</EM>, PUTs and POSTs).
What it does is read any data sent by the client for
a certain time after the server closes the connection. The exact
reasons for doing this are somewhat complicated, but involve what
happens if the client is making a request at the same time the
server sends a response and closes the connection. Without lingering,
the client might be forced to reset its TCP input buffer before it
has a chance to read the server's response, and thus understand why
the connection has closed.
See the <A HREF="#appendix">appendix</A> for more details.<P>
The code in <CODE>lingering_close()</CODE> appears to cause problems
for a number of factors, including the change in traffic patterns
that it causes. The code has been thoroughly reviewed and we are
not aware of any bugs in it. It is possible that there is some
problem in the BSD TCP stack, aside from the lack of a timeout
for the FIN_WAIT_2 state, exposed by the <CODE>lingering_close</CODE>
code that causes the observed problems.<P>
<H2><LI>What can I do about it?</H2>
There are several possible workarounds to the problem, some of
which work better than others.<P>
<H3>Add a timeout for FIN_WAIT_2</H3>
The obvious workaround is to simply have a timeout for the FIN_WAIT_2 state.
This is not specified by the RFC, and could be claimed to be a
violation of the RFC, but it is widely recognized as being necessary.
The following systems are known to have a timeout:
<P>
<UL>
<LI><A HREF="http://www.freebsd.org/">FreeBSD</A> versions starting at
2.0 or possibly earlier.
<LI><A HREF="http://www.netbsd.org/">NetBSD</A> version 1.2(?)
<LI><A HREF="http://www.openbsd.org/">OpenBSD</A> all versions(?)
<LI><A HREF="http://www.bsdi.com/">BSD/OS</A> 2.1, with the
<A HREF="ftp://ftp.bsdi.com/bsdi/patches/patches-2.1/K210-027">
K210-027</A> patch installed.
<LI><A HREF="http://www.sun.com/">Solaris</A> as of around version
2.2. The timeout can be tuned by using <CODE>ndd</CODE> to
modify <CODE>tcp_fin_wait_2_flush_interval</CODE>, but the
default should be appropriate for most servers and improper
tuning can have negative impacts.
<LI><A HREF="http://www.linux.org/">Linux</A> 2.0.x and
earlier(?)
<LI><A HREF="http://www.hp.com/">HP-UX</A> 10.x defaults to
terminating connections in the FIN_WAIT_2 state after the
normal keepalive timeouts. This does not
refer to the persistent connection or HTTP keepalive
timeouts, but the <CODE>SO_LINGER</CODE> socket option
which is enabled by Apache. This parameter can be adjusted
by using <CODE>nettune</CODE> to modify parameters such as
<CODE>tcp_keepstart</CODE> and <CODE>tcp_keepstop</CODE>.
In later revisions, there is an explicit timer for
connections in FIN_WAIT_2 that can be modified; contact HP
support for details.
<LI><A HREF="http://www.sgi.com/">SGI IRIX</A> can be patched to
support a timeout. For IRIX 5.3, 6.2, and 6.3,
use patches 1654, 1703 and 1778 respectively. If you
have trouble locating these patches, please contact your
SGI support channel for help.
<LI><A HREF="http://www.ncr.com/">NCR's MP RAS Unix</A> 2.xx and
3.xx both have FIN_WAIT_2 timeouts. In 2.xx it is non-tunable
at 600 seconds, while in 3.xx it defaults to 600 seconds and
is calculated based on the tunable "max keep alive probes"
(default of 8) multiplied by the "keep alive interval" (default
75 seconds).
<LI><A HREF="http://www.sequent.com">Sequent's ptx/TCP/IP for
DYNIX/ptx</A> has had a FIN_WAIT_2 timeout since around
release 4.1 in mid-1994.
</UL>
<P>
The following systems are known to not have a timeout:
<P>
<UL>
<LI><A HREF="http://www.sun.com/">SunOS 4.x</A> does not and
almost certainly never will have one because it as at the
very end of its development cycle for Sun. If you have kernel
source should be easy to patch.
</UL>
<P>
There is a
<A HREF="http://www.apache.org/dist/contrib/patches/1.2/fin_wait_2.patch"
>patch available</A> for adding a timeout to the FIN_WAIT_2 state; it
was originally intended for BSD/OS, but should be adaptable to most
systems using BSD networking code. You need kernel source code to be
able to use it. If you do adapt it to work for any other systems,
please drop me a note at <A HREF="mailto:marc@apache.org">marc@apache.org</A>.
<P>
<H3>Compile without using <CODE>lingering_close()</CODE></H3>
It is possible to compile Apache 1.2 without using the
<CODE>lingering_close()</CODE> function. This will result in that
section of code being similar to that which was in 1.1. If you do
this, be aware that it can cause problems with PUTs, POSTs and
persistent connections, especially if the client uses pipelining.
That said, it is no worse than on 1.1, and we understand that keeping your
server running is quite important.<P>
To compile without the <CODE>lingering_close()</CODE> function, add
<CODE>-DNO_LINGCLOSE</CODE> to the end of the
<CODE>EXTRA_CFLAGS</CODE> line in your <CODE>Configuration</CODE> file,
rerun <CODE>Configure</CODE> and rebuild the server.
<P>
<H3>Use <CODE>SO_LINGER</CODE> as an alternative to
<CODE>lingering_close()</CODE></H3>
On most systems, there is an option called <CODE>SO_LINGER</CODE> that
can be set with <CODE>setsockopt(2)</CODE>. It does something very
similar to <CODE>lingering_close()</CODE>, except that it is broken
on many systems so that it causes far more problems than
<CODE>lingering_close</CODE>. On some systems, it could possibly work
better so it may be worth a try if you have no other alternatives. <P>
To try it, add <CODE>-DUSE_SO_LINGER -DNO_LINGCLOSE</CODE> to the end of the
<CODE>EXTRA_CFLAGS</CODE> line in your <CODE>Configuration</CODE>
file, rerun <CODE>Configure</CODE> and rebuild the server. <P>
<STRONG>NOTE:</STRONG> Attempting to use <CODE>SO_LINGER</CODE> and
<CODE>lingering_close()</CODE> at the same time is very likely to do
very bad things, so don't.<P>
<H3>Increase the amount of memory used for storing connection state</H3>
<DL>
<DT>BSD based networking code:
<DD>BSD stores network data, such as connection states,
in something called an mbuf. When you get so many connections
that the kernel does not have enough mbufs to put them all in, your
kernel will likely crash. You can reduce the effects of the problem
by increasing the number of mbufs that are available; this will not
prevent the problem, it will just make the server go longer before
crashing.<P>
The exact way to increase them may depend on your OS; look
for some reference to the number of "mbufs" or "mbuf clusters". On
many systems, this can be done by adding the line
<CODE>NMBCLUSTERS="n"</CODE>, where <CODE>n</CODE> is the number of
mbuf clusters you want to your kernel config file and rebuilding your
kernel.<P>
</DL>
<H3>Disable KeepAlive</H3>
<P>If you are unable to do any of the above then you should, as a last
resort, disable KeepAlive. Edit your httpd.conf and change "KeepAlive On"
to "KeepAlive Off".
<H2><LI>Feedback</H2>
If you have any information to add to this page, please contact me at
<A HREF="mailto:marc@apache.org">marc@apache.org</A>.<P>
<H2><A NAME="appendix"><LI>Appendix</A></H2>
<P>
Below is a message from Roy Fielding, one of the authors of HTTP/1.1.
<H3>Why the lingering close functionality is necessary with HTTP</H3>
The need for a server to linger on a socket after a close is noted a couple
times in the HTTP specs, but not explained. This explanation is based on
discussions between myself, Henrik Frystyk, Robert S. Thau, Dave Raggett,
and John C. Mallery in the hallways of MIT while I was at W3C.<P>
If a server closes the input side of the connection while the client
is sending data (or is planning to send data), then the server's TCP
stack will signal an RST (reset) back to the client. Upon
receipt of the RST, the client will flush its own incoming TCP buffer
back to the un-ACKed packet indicated by the RST packet argument.
If the server has sent a message, usually an error response, to the
client just before the close, and the client receives the RST packet
before its application code has read the error message from its incoming
TCP buffer and before the server has received the ACK sent by the client
upon receipt of that buffer, then the RST will flush the error message
before the client application has a chance to see it. The result is
that the client is left thinking that the connection failed for no
apparent reason.<P>
There are two conditions under which this is likely to occur:
<OL>
<LI>sending POST or PUT data without proper authorization
<LI>sending multiple requests before each response (pipelining)
and one of the middle requests resulting in an error or
other break-the-connection result.
</OL>
<P>
The solution in all cases is to send the response, close only the
write half of the connection (what shutdown is supposed to do), and
continue reading on the socket until it is either closed by the
client (signifying it has finally read the response) or a timeout occurs.
That is what the kernel is supposed to do if SO_LINGER is set.
Unfortunately, SO_LINGER has no effect on some systems; on some other
systems, it does not have its own timeout and thus the TCP memory
segments just pile-up until the next reboot (planned or not).<P>
Please note that simply removing the linger code will not solve the
problem -- it only moves it to a different and much harder one to detect.
</OL>
<!--#include virtual="footer.html" -->
</BODY>
</HTML>
|
