INN 2.5.4 strange crash

[Petr Novopashenniy]

On Tue, 3 Feb 2015, Julien ?LIE wrote:

[dd]

J?? > I saw "blocked sleeping" before (for my two channels, innfeed-neva1 and
J?? > innfeed-neva2 at varios times), but not in the time of crash.
J?? 
J?? They do not seem related then.
J?? 

May be..

[dd]

J?? When you have such logs, please paste them so as to check the closing is
J?? fine.
J?? 

I looked more closely my logs, and this is partial information from them:

Feb  4 08:34:33 ns innd: innfeed-neva1!:26:proc:15113 cant write: Resource 
temporarily unavailable
Feb  4 08:34:33 ns innd: innfeed-neva1!:26:proc:15113 blocked sleeping 5
[dd]
Feb  4 08:34:34 ns innd: innfeed-neva1!:26:proc:15113 cant write: Resource 
temporarily unavailable
Feb  4 08:34:34 ns innd: innfeed-neva1!:26:proc:15113 blocked sleeping 5
[dd]
Feb  4 08:34:39 ns innd: innfeed-neva1!:26:proc:15113 wakeup
Feb  4 08:34:39 ns last message repeated 33 times
[dd]
Feb  4 08:34:39 ns innd: innfeed-neva1!:26:proc:15113 wakeup
Feb  4 08:34:39 ns last message repeated 85 times

$ cat news.notice |grep " wakeup"|wc -l
   11103
$

And EAGAIN only twice.

No "sleeping without Waker" messages.

I not see messages with "wakeup" and EAGAIN (temporarily unavailable) in 
errlog and news.err, it surprises me.

Continued logs:

We receive control message:

Feb  4 10:32:46 ns innd: innfeed-neva1!:26:proc:15113 wakeup
Feb  4 10:32:46 ns last message repeated 55 times
Feb  4 10:32:46 ns controlchan[15112]: control_newgroup, 
alt.binaries.history 153fishare at gmail.com 153fishare at gmail.com 
@03006E657773310000000003051400065DD3
@, , , UTF-8, neva-out.feeder.erje.net, doit, 1 
Feb  4 10:32:46 ns innd: ctlinnd command 
k:alt.binaries.history:y:153fishare at gmail.com
Feb  4 10:32:47 ns innd: controlchan! closed
Feb  4 10:32:47 ns innd: innfeed-neva1! closed
Feb  4 10:32:47 ns innfeed[15113]: ME source lost . Exiting

[dd]

Feb  4 10:33:08 ns innd: controlchan! exit 0 elapsed 14017 pid 15112
Feb  4 10:33:08 ns innd: controlchan!:26:proc:26571 26 sleeping without 
Waker
Feb  4 10:33:08 ns innd: controlchan!:26:proc:26571 closed
Feb  4 10:33:08 ns innd: controlchan! exit 0 elapsed 0 pid 26571
Feb  4 10:33:08 ns innd: controlchan!:99:file closed
Feb  4 10:33:09 ns innd: controlchan! spawned controlchan!:99:proc:26578
Feb  4 10:33:09 ns innd: controlchan! restarted

[dd]

Feb  4 10:33:09 ns innd: goblin1.stu.neva.ru:26 26 sleeping without Waker
Feb  4 10:33:09 ns innd: goblin1.stu.neva.ru:26 checkpoint seconds 0 
accepted 0 refused 0 rejected 0 duplicate 0 accepted size 0 duplicate size 
0 rejected size 0
Feb  4 10:33:09 ns innd: goblin1.stu.neva.ru:26 closed seconds 0 accepted 
0 refused 0 rejected 0 duplicate 0 accepted size 0 duplicate size 0 
rejected size 0
Feb  4 10:33:09 ns innd: goblin3.stu.neva.ru:26 26 sleeping without Waker
Feb  4 10:33:09 ns innd: goblin3.stu.neva.ru:26 checkpoint seconds 0 
accepted 0 refused 0 rejected 0 duplicate 0 accepted size 0 duplicate size 
0 rejected size 0
Feb  4 10:33:09 ns innd: goblin3.stu.neva.ru:26 closed seconds 0 accepted 
0 refused 0 rejected 0 duplicate 0 accepted size 0 duplicate size 0 
rejected size 0

[and same for several peers]

Feb  4 10:33:10 ns innd: free:-1 26 free but was in SMASK

And no errors after that.
Looks like problem with my system, I see same fd 26.

In my old logs, I see different fd numbers, but problem is same.

--pety