All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Brian J. Murrell" <brian-SquOHqY54CVWr29BmMi2cA@public.gmane.org>
To: linux-nfs@vger.kernel.org
Subject: Re: stuck/hung nfsv4 mounts
Date: Mon, 03 Nov 2008 12:50:31 -0500	[thread overview]
Message-ID: <1225734631.2247.76.camel@brian-laptop> (raw)
In-Reply-To: <1225731544.6958.6.camel-rJ7iovZKK19ZJLDQqaL3InhyD016LWXt@public.gmane.org>

On Mon, 2008-11-03 at 11:59 -0500, Trond Myklebust wrote:
> 
> BTW: NFSv4 + soft == BAD BAD BAD!

Ahh.  OK.

> You will see weird behaviour, as the
> client and the server will have divergent opinions on what happened
> if/when an RPC call fails.

I have removed the soft option from the auto map and killed the
in-progress mount.nfs4.  Seems the new one is not using soft any more:

root     22731 22730  0 12:06 ?        00:00:00 /sbin/mount.nfs4 linux:/home/share /autohome/share -s -o rw,sec=krb5i,intr,rsize=8192,wsize=8192

> Otherwise, have you checked on the state of your rpc.gssd? It looked as
> if several of those traces were waiting around RPCSEC_GSS upcalls...
I thought it was working.  I killed it and restarted it with -vvv -rrr
and this is what it said when automount re-issued the above mount:

Nov  3 12:02:15 pc rpc.gssd[21773]: rpcsec_gss: debug level is 3
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_create_default()
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_create()
Nov  3 12:06:00 pc rpc.gssd[21774]: authgss_create: name is 0x8ab8d88
Nov  3 12:06:00 pc rpc.gssd[21774]: authgss_create: gd->name is 0x8ab8ed0
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_refresh()
Nov  3 12:06:00 pc rpc.gssd[21774]: struct rpc_gss_sec: 
Nov  3 12:06:00 pc rpc.gssd[21774]:      mechanism_OID: { 1 2 134 72 134 247 18 1 2 2 } 
Nov  3 12:06:00 pc rpc.gssd[21774]:      qop: 0 
Nov  3 12:06:00 pc rpc.gssd[21774]:      service: 1 
Nov  3 12:06:00 pc rpc.gssd[21774]:      cred: 0x8abb830 
Nov  3 12:06:00 pc rpc.gssd[21774]:      req_flags: 00000002 
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_marshal()
Nov  3 12:06:00 pc rpc.gssd[21774]: xdr_rpc_gss_buf: encode success ((nil):0)
Nov  3 12:06:00 pc rpc.gssd[21774]: xdr_rpc_gss_cred: encode success (v 1, proc 1, seq 0, svc 1, ctx (nil):0)
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_wrap()
Nov  3 12:06:00 pc rpc.gssd[21774]: xdr_rpc_gss_buf: encode success (0x8abbeb8:483)
Nov  3 12:06:00 pc rpc.gssd[21774]: xdr_rpc_gss_init_args: encode success (token 0x8abbeb8:483)
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_validate()
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_unwrap()
Nov  3 12:06:00 pc rpc.gssd[21774]: xdr_rpc_gss_buf: decode success (0x8abbe98:4)
Nov  3 12:06:00 pc rpc.gssd[21774]: xdr_rpc_gss_buf: decode success (0x8abc110:114)
Nov  3 12:06:00 pc rpc.gssd[21774]: xdr_rpc_gss_init_res decode success (ctx 0x8abbe98:4, maj 0, min 0, win 128, token 0x8abc110:114)
Nov  3 12:06:00 pc rpc.gssd[21774]: authgss_create_default: freeing name 0x8ab8d88
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_get_private_data()
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_free_private_data()
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_destroy()
Nov  3 12:06:00 pc rpc.gssd[21774]: in authgss_destroy_context()
Nov  3 12:06:00 pc rpc.gssd[21774]: authgss_destroy: freeing name 0x8ab8ed0

The kdc logged:

Nov  3 12:06:00 linux krb5kdc[5006]: AS_REQ (1 etypes {1}) 10.75.22.1: ISSUE: authtime 1225731960, etypes {rep=1 tkt=16 ses=1}, nfs/pc.interlinx.bc.ca@ILINX for krbtgt/ILINX@ILINX
Nov  3 12:06:00 linux krb5kdc[5006]: TGS_REQ (1 etypes {1}) 10.75.22.1: ISSUE: authtime 1225731960, etypes {rep=1 tkt=1 ses=1}, nfs/pc.interlinx.bc.ca@ILINX for nfs/linux.interlinx.bc.ca@ILINX

in correlation to the new mount request, but the mount.nfs4 didn't
complete.

b.



  parent reply	other threads:[~2008-11-03 17:50 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-11-03 15:05 stuck/hung nfsv4 mounts Brian J. Murrell
2008-11-03 16:59 ` Trond Myklebust
     [not found]   ` <1225731544.6958.6.camel-rJ7iovZKK19ZJLDQqaL3InhyD016LWXt@public.gmane.org>
2008-11-03 17:25     ` Jim Rees
2008-11-03 17:37       ` Trond Myklebust
     [not found]         ` <1225733834.6958.12.camel-rJ7iovZKK19ZJLDQqaL3InhyD016LWXt@public.gmane.org>
2008-11-03 21:40           ` Chuck Lever
2008-11-03 22:20             ` Peter Staubach
2008-11-03 22:47               ` Chuck Lever
2008-11-04 16:03                 ` Peter Staubach
2008-11-03 17:38       ` Benny Halevy
2008-11-03 17:50     ` Brian J. Murrell [this message]
2008-11-03 19:58       ` Kevin Coffman
     [not found]         ` <4d569c330811031158r26963e0w5bcf8331e0fb14b7-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2008-11-03 20:09           ` Brian J. Murrell
2008-11-03 20:28             ` Kevin Coffman
     [not found]               ` <4d569c330811031228r5bb9aefs7a970303910810e2-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2008-11-03 21:12                 ` Brian J. Murrell
2008-11-03 22:33                   ` Kevin Coffman
     [not found]                     ` <4d569c330811031433k7ae18d4enfbda349e8f90a951-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2008-11-03 23:45                       ` Brian J. Murrell
2008-11-04  1:24                       ` Brian J. Murrell
2008-11-04 15:14                         ` Brian J. Murrell
2008-11-04 17:22                         ` Kevin Coffman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1225734631.2247.76.camel@brian-laptop \
    --to=brian-squohqy54cvwr29bmmi2ca@public.gmane.org \
    --cc=linux-nfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.