t/lru-crawler.t sometimes fail

[GoogleCodeExporter]

What steps will reproduce the problem?
1. build memcached
2. run "make test"

What is the expected output?
End of testsuite witout any error

What do you see instead?
testsuite sometimes stop with:
 Timeout.. killing the process
 t/lru-crawler.t ...... 
 Failed 126/221 subtests 

What version of the product are you using? On what operating system?
1.4.24 on Debian GNU/Linux unstable

Please provide any additional information below.

Other users seems to reproduce this randomly:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=789835
http://forums.famillecollet.com/viewtopic.php?id=3165
https://code.google.com/p/memcached/issues/detail?id=398
http://webcache.googleusercontent.com/search?q=cache:2j2npL8eOAMJ:https://arch-c
i.org/extra/memcached/log/+&cd=10&hl=fr&ct=clnk

Original issue reported on code.google.com by [email protected] on 29 Jun 2015 at 8:33

[GoogleCodeExporter]

how often does this happen? Does it happen if you run the test on its own in a 
loop? like:
while true ; do prove t/lru-crawler.t ; done

Original comment by [email protected] on 5 Jul 2015 at 1:09

[GoogleCodeExporter]

I ran the test in a loop (~ 248 iterations) and does not reproduce the problem 
at all.

Just 2 iterations fail with:

failed to listen on TCP port 45378: Address already in use
Failed to startup/connect to memcached server. at t/lru-crawler.t line 79.
# Looks like you planned 221 tests but ran 189.
# Looks like your test exited with 111 just after 189.
t/lru-crawler.t .. 
Dubious, test returned 111 (wstat 28416, 0x6f00)
Failed 32/221 subtests 

Test Summary Report
-------------------
t/lru-crawler.t (Wstat: 28416 Tests: 189 Failed: 0)
  Non-zero exit status: 111
  Parse errors: Bad plan.  You planned 221 tests but ran 189.
Files=1, Tests=189,  5 wallclock secs ( 0.06 usr  0.00 sys +  0.10 cusr  0.02 
csys =  0.18 CPU)
Result: FAIL

Which is not the same issue; this is really weird because i don't have any 
other process listening on these 2 ports on this box.

This seems to be the same bug as 
https://code.google.com/p/memcached/issues/detail?id=398, I didn't realize that 
i've pointed 

Original comment by [email protected] on 16 Jul 2015 at 12:44