Date: Tue, 06 Apr 1999 17:14:11 -0400 From: "David E. Cross" <crossd@cs.rpi.edu> To: freebsd-hackers@freebsd.org Subject: NFS pause in 3.1-STABLE Message-ID: <199904062114.RAA10003@cs.rpi.edu>
next in thread | raw e-mail | index | archive | help
We had one of our machines "pause" earlier today. The machine is running 3.1-STABLE from March 26. The machine was running fine for 4+ days (has been very stable), when it stoped respondning. It would still answer pings, and new connects would be accept()ed, but apparently never passed onto the process. Existing processes all hung (I had a number of ssh X connections going through the machine, and with no additional disk accesses they all stoped.). From prior experience when this happens, it will eventually recover (anywhere from 45 minuts to hours). We were able to do a tcpdump on the network, and wathced the traffic. What we saw was repeated NFS lookups (with a reply from the server), and then get_attr requests. This was not a spam condition, the requests were only 3 or 3 per second. We identified which file it was asking for and moved it arround until it generated a STALE_NFS error. After that we placed a same named file back and the next lookup found it. Still the machine would not respond. The machine would also not respond to its serial console. Hitting return where a login prompt should have been produced no echo or other response. Any ideas what causes this? It is much rarer than the other, extinct, problems with NFS/TCP and KVA, but still annoying. -- David Cross To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199904062114.RAA10003>