Ubuntu
unison package

unhandled exception with

Bug #839 reported by Robert Collins
28
Affects Status Importance Assigned to Milestone
unison (Ubuntu)
Invalid
Medium
Unassigned

Bug Description

I setup a unison profile to syncronise a tree with 260K files spread over 14K directories

I setup a unison profile to syncronise a tree with 260K files spread over 14K directories. Unison crashes:
It warns me at the start that this is a new sync and everything will be conflicts that isn't already identical.
It then scans for changes for ~ 20 minutes.
Then it stops:
"Uncaught exception Out of memory"
Whats annoying here is that the gui stays running. Also, it should be able to do a partial list or some such, but that would probably be a bigger change ;|.

See original description
Daniel Holbach (dholbach)
Changed in unison:
assignee: nobody → motu
Revision history for this message
Hervé Cauwelier (hcauwelier-deactivatedaccount) wrote :

Please could you try with Breezy (Ubuntu 5.10) when you update to it?

Revision history for this message
Yann Rouillard (yann-pleiades) wrote :

So did you have time to test with breezy ?

Changed in unison:
status: Unconfirmed → Needs Info
Revision history for this message
Robert Collins (lifeless) wrote :

yes and dapper. Still a bug.

Revision history for this message
Emmet Hikory (persia) wrote :

Submitter tested against Breezy and recent Dapper.

Changed in unison:
status: Needs Info → Confirmed
Revision history for this message
Loïc Minier (lool) wrote :

I suppose this is by design, and this is covered in the FAQ:
http://alliance.seas.upenn.edu/~bcpierce/unison/wiki/index.php?n=Main.UnisonFAQTroubleshooting

I tried raising my stack size limit and number of open file descriptor limit to insane values, and unison didn't seem to work any longer.

Stracing the server part of the process shows:
(many read()s)
read(4, "ix) with ESMTP id 93618C2E6;\r\n\tT"..., 7700) = 7700
read(4, ")\r\n\tby mx0.bpl-group.org (Postfi"..., 7700) = 5049
read(4, "", 2651) = 0
mmap2(NULL, 253952, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb69ec000
...
mmap2(NULL, 135168, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb478b000
mremap(0xb478b000, 135168, 266240, MREMAP_MAYMOVE) = 0xb474a000
mremap(0xb474a000, 266240, 528384, MREMAP_MAYMOVE) = 0xb46c9000
mremap(0xb46c9000, 528384, 1052672, MREMAP_MAYMOVE) = 0xb45c8000
mremap(0xb45c8000, 1052672, 2101248, MREMAP_MAYMOVE) = 0xb43c7000
mremap(0xb43c7000, 2101248, 4198400, MREMAP_MAYMOVE) = 0xb3fc6000
mremap(0xb3fc6000, 4198400, 8392704, MREMAP_MAYMOVE) = 0xb37c5000
mremap(0xb37c5000, 8392704, 16781312, MREMAP_MAYMOVE) = 0xb27c4000
mremap(0xb27c4000, 16781312, 33558528, MREMAP_MAYMOVE) = 0xb07c3000
write(2, "Uncaught exception Out of memory"..., 33) = 33
exit_group(1) = ?

It's absolutely not clear why it fails, even with ulimit -a:
-t: cpu time (seconds) unlimited
-f: file size (blocks) unlimited
-d: data seg size (kbytes) unlimited
-s: stack size (kbytes) 262144
-c: core file size (blocks) 0
-m: resident set size (kbytes) unlimited
-u: processes unlimited
-n: file descriptors 8192
-l: locked-in-memory size (kb) unlimited
-v: address space (kb) unlimited
-x: file locks unlimited

I'm trying a ltrace(), but it takes too long with ltracing to reach the bug.

Revision history for this message
Loïc Minier (lool) wrote :

I tried a "-debug all" run (I've given up on ltrace) and got:
[server: remote_emit] Sending result (id: 169)

just before the OOM error.

Revision history for this message
William Grant (wgrant) wrote :

It has been six months, and we've almost got two further releases. Is it present in Feisty, or even Edgy?

Changed in unison:
assignee: motu → nobody
status: Confirmed → Needs Info
Revision history for this message
Mark Reitblatt (mark-reitblatt) wrote :

Closing due to lack of response. If this problem still exists in the latest version of Ubuntu, please reopen.

Changed in unison:
status: Needs Info → Rejected
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.