So presumably one of the descriptors in readfds (a copy of SVC_FDSET) is bad. So it sounds like the select() in my_svc_run is returning EBADF. Is this something that started happening with some update, or that only affects some (e.g., only F19) clients? Utils/mountd/svc_run.c:113: xlog(L_ERROR, "my_svc_run() - select: %m") Utils/statd/svc_run.c:119: xlog(L_ERROR, "my_svc_run() - select: %m") I have hopefully "fixed" the issue by rebuilding nfs-utils from the SRPM (commented-out these lines in the source):
Additional comment from Marek on 07:56:24 EDT. This laptop is also an NFS client, NOT a server. I have had this happen on F19 x86_64 Linux as well (3 times so far).
Additional comment from Steve on 14:19:26 EDT. I "solved" it by 'chmod 000 /usr/sbin/rpc.statd' but I can't use NFS anymore. How reproducible: intermittent, I can't reproduce it at willĪctual results: /var at 100%, the machine either doesn't start or when already started, becomes sluggish and unresponsive. Version-Release number of selected component (if applicable): nfs-utils-1.2.8-3.0.fc19.x86_64 I have /var as a separate partition I thought 8GB was enough for a home PC. +++ This bug was initially created as a clone of Bug #995861 +++ĭescription of problem: /var/log/messages running into GBs in size caused by rpc.statd.
0 kommentar(er)
