[Bug 5478] rsync: writefd_unbuffered failed to write 4092 bytes [sender]: Broken pipe (32)

[samba-bugs at samba.org]

https://bugzilla.samba.org/show_bug.cgi?id=5478

--- Comment #9 from Eric Shubert <ejs at shubes.net> 2011-09-20 14:42:37 UTC ---
(In reply to comment #8)
> By any chance do you consider these files "big" files?  I am asking
> because I saw similar timeout when sending big files (bug7757,
> bug7195).

Not really. Most problem files appear to be 20-30MB, but a few are <10M and
there's one which is only 25K. I'll keep this in mind though.

> Also in your next strace output, could you use "-tt" to
> show timestamps?  Thanks.

Here's one with timestamps:
1636  07:15:20.138388 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
1636  07:15:20.138646 read(0, "\374\17\0\7", 4) = 4
1636  07:15:20.138883 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
1636  07:15:20.139121 read(0,
"\211\313Wq\361\315ogBR\371\346\351\327\335\r@\20\373\276\303\372<F\373\222\333\340\353\272\261\3"...,
4092) = 4092
1636  07:15:20.139466 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
1636  07:15:20.139714 read(0, "\374\17\0\7", 4) = 4
1636  07:15:20.139956 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
1636  07:15:20.140195 read(0,
"\307\371e\245(K\17\351\201O\357\2139\221\322%\345)\201.\252G\364\221\326\256\306u(\2353\346"...,
4092) = 4092
1636  07:15:20.140475 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
1636  07:15:20.140716 read(0, "\374\17\0\7", 4) = 4
1636  07:15:20.140943 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
1636  07:15:20.141180 read(0,
"\221\367\307\26yg0\31\t\314\250Gr\366e\301\231]\211~\253F-\320\240T\177T.\201\206\341"...,
4092) = 1223
1636  07:15:20.141502 select(1, [0], [], NULL, {60, 0} <unfinished ...>
1635  07:15:51.190750 <... select resumed> ) = 0 (Timeout)
1635  07:15:51.190975 select(4, [3], [], NULL, {60, 0} <unfinished ...>
1636  07:16:20.139240 <... select resumed> ) = 0 (Timeout)
1636  07:16:20.139459 select(1, [0], [], NULL, {60, 0} <unfinished ...>
1635  07:16:51.194224 <... select resumed> ) = 0 (Timeout)
1635  07:16:51.194447 select(4, [3], [], NULL, {60, 0} <unfinished ...>

Thanks for having a look at this Tim.

I'm going to put a problem/test file on the target machine's hdd and continue
testing with a local2local transfer. I should have mentioned that the target in
this case is a pair of USB2 HDDs in a raid-1 configuration. I seem to recall
another posting of this error involved writing to a USB drive as well. I expect
at this point that USB might be a contributing factor. I'm stumped though as to
why it has worked fine for over year before coming across this problem.

-- 
Configure bugmail: https://bugzilla.samba.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.