[Bucardo-general] Problems....

[Michelle Sullivan]

Greg Sabino Mullane wrote:
>> Where can I look for the cause of this one please?
>>
>> [Tue Aug  3 14:41:56 2010] CTL Rows updated child 80882 to aborted in q: 1
>> [Tue Aug  3 14:41:56 2010] CTL Warning! Kid 80882 seems to have died.
>> Sync "sorbs2_swap"
> 
> What do the logs show earlier for when this KID died? Also 
> check your Postgres logs for anything of interest.
> 
> 


Nevermind - found it...  not a Bucardo problem, an OS problem (hence no
log)....

Aug  3 14:41:49 stinger kernel: swap zone exhausted, increase kern.maxswzone
Aug  3 14:41:49 stinger kernel: pid 80882 (perl5.10.1), uid 0, was
killed: out of swap space
Aug  3 14:41:49 stinger kernel: swap zone exhausted, increase kern.maxswzone
Aug  3 14:41:49 stinger kernel: pid 80882 (perl5.10.1), uid 0, was
killed: out of swap space
Aug  3 14:41:49 stinger kernel: swap zone exhausted, increase kern.maxswzone
Aug  3 14:41:49 stinger kernel: pid 80882 (perl5.10.1), uid 0, was
killed: out of swap space


Seems the problem is a kernel tunable....  There is 32G of swap
available, only 16G (approx) usable by any single process.. :-/

Michelle