Linux NFS fsync problems

From: byaarov@xxxxxxxxx
Date: Wed, 30 Apr 2008 13:18:10 -0700 (PDT)

Hi,
I have an NFS mount from a linux client to a linux NFS server using a
soft mount.

I have a program that is writing to a file on that share, correctly
checking for return values and after each write calls fsync. It also
logs each message with a unique increasing transaction ID.

When I reboot my NFS server, my client application does not see any
error and fsync continues to return success. When the NFS server
comes back up, the writes keep going through, but all the writes that
were issued while the NFS server was down is lost.

How do I detect that the writes are actually NOT succeeding?

.

Follow-Ups:
- Re: Linux NFS fsync problems
  - From: h.stroph

Prev by Date: Re: Internet connection on Linux
Next by Date: Re: Linux NFS fsync problems
Previous by thread: high-availability: LVS vs HAProxy question
Next by thread: Re: Linux NFS fsync problems
Index(es):
- Date
- Thread

Relevant Pages

nfs connection: mount hangs
... I have a problem connecting a linux client to a nfs server. ... This command never returns and seems to hang, but when I kill it from ... the connection is really established - BUT there is no ...
(comp.os.linux.networking)
Re: Cant edit text -
... word processor is nearly useless because I can't edit anything? ... Are the files on the local machine, or are they on an NFS mount? ... I don't remember how I copied them but I had all the old F-10 files on the nfs server and quite likely took them from there. ...
(Fedora)
Use of automounter for permanent mounts?
... One of my customers has a Web-based application on Solaris 9 with ... This results in processes that are using the NFS mount ... hanging until the NFS server comes back. ... that the failed mount has other side-effects such as a delay in ...
(comp.unix.solaris)
Re: Delete question
... > In my NFS mount I have .Trash-501 which is where the files go when I ... > .Trash-501/files directory is empty. ... > to the trash when they are deleted. ... the NFS server for user 501. ...
(alt.os.linux.suse)
Re: F12 NFS Failures
... This was using md5sum for about 10GB over the NFS mount ... The default for F12/Centos5.4 appears to be TCP - which freezes ... Forcing UDP gives NO errors for 10GB transfer ... I know this is an old thread, but I thought I'd toss in that you will see symptoms very much like this if only one of your machines (probably the NFS server) is configured to use jumbo frames. ...
(Fedora)