Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.bugs.dist > #952285

Bug#926046: Negotiated or default wsize causes misbehavior

Cross-posted to 2 groups.

Show all headers | View raw

Package: nfs-common
Version: 1:1.3.4-2.1

I'm using NFSv4 over TCP at the moment.  If I don't specify rsize and
wsize on the client, either the client negotiates a wsize of 256KB or
defaults to a wsize of 256KB ("wsize=262144").

When dumping large amounts of data (moving 2TB of data around, figure
many 200MB files) onto the server, after a while the mount hangs and then
messages start appearing in the server kernel log:
"[sss.mmm] NFSD: client xxxxx testing state ID with incorrect client ID"
After several minutes the mount was recovering, but having an entire
machine locked up for a while is a problem.

During an attempt to revert to using UDP, I discovered that explicitly
setting wsize=8192 fixed the problem (this size is reasonable with UDP if
you've got jumbo-frame support).  I'm guessing either the default is bad
or negotiation is failing to generate a working value.


-- 
(\___(\___(\______          --=> 8-) EHM <=--          ______/)___/)___/)
 \BS (    |         ehem+sigmsg@m5p.com  PGP 87145445         |    )   /
  \_CS\   |  _____  -O #include <stddisclaimer.h> O-   _____  |   /  _/
8A19\___\_|_/58D2 7E3D DDF4 7BA6 <-PGP-> 41D1 B375 37D0 8714\_|_/___/5445

Back to linux.debian.bugs.dist | Previous | Next | Find similar | Unroll thread

Thread

Bug#926046: Negotiated or default wsize causes misbehavior Elliott Mitchell <ehem+debian@m5p.com> - 2019-03-30 21:30 +0100

csiph-web