Date: Tue, 24 Jan 2006 17:22:15 +0100 From: Cedric Tabary <tabary@efrei.fr> To: freebsd-stable@freebsd.org Subject: Re: nfs_getpages: error 70 vm_fault: pager read error Message-ID: <20060124162215.GF4416@efrei.fr> In-Reply-To: <20060124141510.GA68870@xor.obsecurity.org> References: <20060124082301.GB4416@efrei.fr> <20060124141510.GA68870@xor.obsecurity.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 24/01/2006 09:15, Kris Kennaway wrote: > On Tue, Jan 24, 2006 at 09:23:01AM +0100, Cedric Tabary wrote: > > I have 6 dell 1850, dual xeon, 2G ram, no HTT > > 4 are running RELENG_6 from Mon Jan 9 > > 2 are running RELENG_5_4 from Tue Oct 4 > > > > They are used as web servers apache 1.3 > > web pages are on a NetAPP F760 with nfs options : > > rw,soft,intr,nfsv3,mntudp,-r=8192,-w=8192,noatime,acregmin=30 > > > > the dmesg is full of : > > > > pid 34397 (httpd), uid 100: exited on signal 11 > > pid 80101 (httpd), uid 100: exited on signal 6 > > nfs_getpages: error 70 > > vm_fault: pager read error, pid 78999 (httpd) > > pid 78999 (httpd), uid 100: exited on signal 11 > > pid 15738 (httpd), uid 100: exited on signal 11 > > pid 14910 (httpd), uid 100: exited on signal 11 > > nfs_getpages: error 70 > > vm_fault: pager read error, pid 15981 (httpd) > > pid 15981 (httpd), uid 100: exited on signal 11 > > > > pid 84481 (httpd), uid 100: exited on signal 11 > > nfs_getpages: error 60 > > vm_fault: pager read error, pid 54468 (httpd) > > pid 46271 (httpd), uid 100: exited on signal 11 > > > > pid 54916 (httpd), uid 100: exited on signal 11 > > pid 28818 (httpd), uid 100: exited on signal 11 > > nfs_getpages: error 5 > > vm_fault: pager read error, pid 93494 (httpd) > > pid 2734 (httpd), uid 100: exited on signal 11 > > > > > > Do you have an idea of what happens ? > > Are you sure httpd itself (or some of the shared libraries, etc) are > not being run over NFS? This error will happen when you change a > binary on the server while it is running on the client. > > Kris Thanks for the clue ! I think I found the problem, i have this sometimes in the apache logs : [15981] EACCELERATOR: PHP crashed on opline 37 of file() at xxxxxxx:66 I have the segfault a few lines later and a nfs_getpages + vm_fault in /var/log/messages at the same time ! It seems like a bug in php or eaccelerator when accessing a file when it is edited by another server... Cédric
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060124162215.GF4416>