> The machine in question has 64M RAM and 128M swap and was handling a
> couple of hundred concurrent connections through a custom proxy. The
> 'unable to load interpreter' message appeared in the log 3 times but the
> box seems to have recovered now (it was using about 70megs of swap when I
> got to it a minute or two later).
The machine I referenced earlier (with the same errors on a 2.0.35 kernel)
has 512M RAM and 128M swap. I'm not sure what exactly it was handling a
the time, but overall, the machine handles nearly 1M hits a day as a web
server.
[snip]
> I can't find any more useful info than that. I think it's unlikely that I
> was out of both swap and RAM so I'd guess the VM balancing is not perfect.
Ditto to the swap/RAM thing; the only way I could imagine totally running
out of swap/RAM would be with a horrible runaway process. Could it be that
nameservice gets confused and name requests queue up too far?
In any case, what can be done to avoid this (besides, in my case, keeping
the name server from crashing)? Thanks for the suggestions!
/pg
-- Peter Green Gospel Communications Network, SysAdmin pcg@gospelcom.net
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/