JF Guilmard writes:...
There is something weird at /var/log/messages when the pb occurs: Jun 8 04:26:33 anarki kernel: nfs: server xxx.xxx.xxx.xxx not responding, still trying
Jun 8 04:26:35 anarki kernel: nfs: server xxx.xxx.xxx.xxx OK
And last but not least, it seems that a privileged user has a very big MailBox ( almost 4giga) and the problem occurs almost at the time when he makes a search in his mailbox. So, do you guys have an idea about what's happen ?
Very likely that your network does not have sufficient capacity to handle such a load.
I'll also point out that you are very likely to see this error when your NFS server is using ext3. ext3 is helluva slow, and the Linux kernel will sometimes just stall for 40-60 seconds while it thinks about what it's going to do to the fs. It sucks. Real bad. When the thing stalls, no local proccesses get CPU time (whether they're reading from that disk or not), and no network IO happens. The box just sits. It happens a lot when you're opening many files or just stat()ing many files in succession.
ext2 doesn't have that problem. The problem becomes much less significant on ext3 if you mount the fs using the "noatime" option. Aside from that, I haven't found a way to make this problem significantly less... uh.. problematic. I'm guessing that there's something wrong with the write-reordering code in ext3. You might be able to further reduce the problem by playing with the bdflush values on the server, making it flush more often, but I haven't found a good set of values yet.
I haven't yet benchmarked XFS, either, but it's in my list of stuff to do.
------------------------------------------------------- This SF.net email is sponsored by: Etnus, makers of TotalView, The best thread debugger on the planet. Designed with thread debugging features you've never dreamed of, try TotalView 6 free at www.etnus.com. _______________________________________________ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
