Re: [Xen-devel] [xen-unstable test] 13944: regressions - FAIL

[Dario Faggioli <raistlin.df@xxxxxxxxx>]

On Tue, 2012-10-09 at 19:00 +0100, Ian Jackson wrote: 
> This is this:
> 
> fatal: Out of memory? mmap failed: Cannot allocate memory
> fatal: The remote end hung up unexpectedly
> make[1]: *** [qemu-xen-dir-find] Error 128
> 
> On gall-mite:
> 
> root@gall-mite:~# free -m
>              total       used       free     shared    buffers     cached
> Mem:          8102       2059       6042          0        650       1286
> -/+ buffers/cache:        122       7980
> Swap:        12143          0      12143
> root@gall-mite:~# 
> 
> So it has 8G of RAM and 12G of swap.
> 
> I wonder if the git object cache is too big.  I will clear gall-mite's
> and see if it helps.
> 
Mmm, gall-mite is the box on which _all_ the sched=sedf test were
running for quite a bit of flights, and consistently failing at the
xen-boot stage. I pointed out the situation already here:

http://www.gossamer-threads.com/lists/xen/devel/258909?do=post_view_threaded

Now that the very same test has moved to some other host it seems to be
working again:

http://www.chiark.greenend.org.uk/~xensrcts/logs/13943/test-amd64-amd64-xl-sedf/info.html

That could be something very silly to think (in which case sorry), but
might this all be some hardware issue on that machine?

Regards,
Dario



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel