I assume you're not on the openafs-devel list. It takes a lot longer for anyone to see list posts from you if you're not on the list, you know...
On Fri, 23 Jul 2010 07:04:03 -0500 Frank Bagehorn <[email protected]> wrote: > I digged into the /var/log/messages and found something: Based on this > I changed the client settings for the memcache. I went from > "AUTOMATIC" to "65536" (64MB) and that worked. Going up to 96MB worked, > 128MB crashed the machine again. Tried 97MB and it crashed as well. 96MB > seems to be the limit. [...] >> Jul 23 13:21:00 testinstall kernel: [ 736.190179] vmap allocation failed - >> use vmalloc=<size> to increase size. >> Jul 23 13:21:00 testinstall kernel: [ 736.190197] vmalloc size=11000 >> start=f77fe000 end=feffe000 node=-1 gfp=d2 Your problem is the same as the one here: <https://lists.openafs.org/pipermail/openafs-info/2010-July/034100.html> For whatever reason x86 linux kernels only allow you to vmalloc about 100M of memory if you have around 1G of RAM or more. Either specify vmalloc=<size> as the message tells you, or use a smaller cache, or use a disk cache. (Or complain to the linux folks and get them to change/fix it) I don't know why the machine locks up, but I'm not sure if that's our fault... in Claudio's case it seems the machine was fine aside from AFS. -- Andrew Deason [email protected] _______________________________________________ OpenAFS-devel mailing list [email protected] https://lists.openafs.org/mailman/listinfo/openafs-devel
