no option to override the hardcoded 30 seconds on socket timeout

[GoogleCodeExporter]

Hi, I created a patch for it, would it be possible to have it included??

this is the effect... having a netcat in listening...

dyn-86-client:apachebench-standalone-read-only zeph$ time ./ab -s 2 
'http://localhost:9000/'
This is ApacheBench, Version 2.3 <$Revision$>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking localhost (be patient)...apr_poll: The timeout specified has 
expired (70007)

real    0m2.006s
user    0m0.001s
sys 0m0.003s

dyn-86-client:apachebench-standalone-read-only zeph$ time ./ab -s 70 
'http://localhost:9000/'
This is ApacheBench, Version 2.3 <$Revision$>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking localhost (be patient)...apr_poll: The timeout specified has 
expired (70007)

real    1m10.005s
user    0m0.001s
sys 0m0.004s

 + + + 

I attached a patch that allows such a behaviour

without this, it is basically impossible to test webservices that might take 
for more that 30 secs per request, especially under high load...

Original issue reported on code.google.com by guido.se...@gmail.com on 25 Jul 2012 at 4:01

Attachments:

• apachebench_socket_timeout.patch

[GoogleCodeExporter]

p.s. also the man file needs to be updated, but it is not available in your 
repo...

Original comment by guido.se...@gmail.com on 25 Jul 2012 at 4:04

[GoogleCodeExporter]

i'd like to see this included, please do.

Original comment by sebastia...@gtempaccount.com on 26 Jul 2012 at 7:42

[GoogleCodeExporter]

How does one go about using the patch? Tried looking for how to do this on the 
wiki, but no luck.

Original comment by lighton.phiri@gmail.com on 15 Dec 2012 at 10:16

[GoogleCodeExporter]

k, I'll see again how I can get full build of the upstream of the Apache 
Project and not this one
...and try to apply for a merge

Original comment by guido.se...@gmail.com on 15 Dec 2012 at 10:23

[GoogleCodeExporter]

k, patches against afs trunk created and submitted... let's keep fingers 
crossed :)

(added also here if the owner of this project will decide to merge them too... 
one day :p )

Original comment by guido.se...@gmail.com on 16 Dec 2012 at 2:30

Attachments:

• ab_man.diff

[GoogleCodeExporter]

uhmm, I added only the man patch, here comes the ab.c patch (adding it)

Original comment by guido.se...@gmail.com on 16 Dec 2012 at 2:31

Attachments:

• ab.diff

[GoogleCodeExporter]

What of the -t argument in ab 2.3? Does that not override the socket timeout?

./ab -h
Usage: ./ab [options] [http[s]://]hostname[:port]/path
Options are:
    -n requests     Number of requests to perform
    -c concurrency  Number of multiple requests to make
    -t timelimit    Seconds to max. wait for responses
etc.

Original comment by scottsti...@gmail.com on 26 Jun 2013 at 3:33

[GoogleCodeExporter]

dunno about ab 2.3, but ab 2.2 shows the same text as above for 

$ ab -h

but the man page is more explicit:

Maximum number of seconds to spend for benchmarking. This implies a -n 50000 
internally. Use this to benchmark the server within a fixed total amount of 
time. Per default  there is no timelimit.

so, no -t is something different (in 2.2) .

Original comment by sebastia...@gtempaccount.com on 27 Jun 2013 at 7:51

[GoogleCodeExporter]

p.s. it is merged as -s ... I don't know in which release will end up in
... certainly someone will have to backport it into this project too

http://mail-archives.apache.org/mod_mbox/httpd-dev/201212.mbox/%3C50CF330D.10103
@wanadoo.fr%3E

Original comment by guido.se...@gmail.com on 27 Jun 2013 at 8:25