linux-nfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Kinglong Mee <kinglongmee@gmail.com>
To: Trond Myklebust <trond.myklebust@primarydata.com>
Cc: Linux NFS Mailing List <linux-nfs@vger.kernel.org>
Subject: [PATCH] NFS4: Retry destroy session when getting -NFS4ERR_DELAY
Date: Fri, 20 Mar 2015 16:31:28 +0800	[thread overview]
Message-ID: <550BDAE0.2070409@gmail.com> (raw)

When umounting a client, it cost near ten seconds.
Dump the request, got
   client                     server
DELEGRETURN              ---->
DESTROY_SESSION          ----> 
          NFS4ERR_DELAY  <----  DESTROY_SESSION reply
                NFS4_OK  <----  DELEGRETURN reply
DESTROY_CLIENTID          ---->
  NFS4ERR_CLIENTID_BUSY  <----  DESTROY_CLIENTID reply
DESTROY_CLIENTID          ---->
  NFS4ERR_CLIENTID_BUSY  <----  DESTROY_CLIENTID reply
         ... ....                  ... ... 
There are ten DESTROY_CLIENTID requests.
This patch retry DESTROY_SESSION when getting NFS4ERR_DELAY,
try the best to destroy the session as destroy clientid.

With this patch, only cost more than 1 seconds, as,
   client                     server
DELEGRETURN          ---->
DESTROY_SESSION      ----> 
     NFS4ERR_DELAY  <----  DESTROY_SESSION reply
           NFS4_OK  <----  DELEGRETURN reply
DESTROY_SESSION      ----> 
           NFS4_OK  <----  DESTROY_SESSION reply
DESTROY_CLIENTID     ---->
           NFS4_OK  <----  DESTROY_CLIENTID reply

Signed-off-by: Kinglong Mee <kinglongmee@gmail.com>
---
 fs/nfs/nfs4proc.c | 26 +++++++++++++++++++++-----
 1 file changed, 21 insertions(+), 5 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 627f37c..2631dc2 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -7320,7 +7320,7 @@ out:
  * Issue the over-the-wire RPC DESTROY_SESSION.
  * The caller must serialize access to this routine.
  */
-int nfs4_proc_destroy_session(struct nfs4_session *session,
+static int _nfs4_proc_destroy_session(struct nfs4_session *session,
 		struct rpc_cred *cred)
 {
 	struct rpc_message msg = {
@@ -7332,10 +7332,6 @@ int nfs4_proc_destroy_session(struct nfs4_session *session,
 
 	dprintk("--> nfs4_proc_destroy_session\n");
 
-	/* session is still being setup */
-	if (!test_and_clear_bit(NFS4_SESSION_ESTABLISHED, &session->session_state))
-		return 0;
-
 	status = rpc_call_sync(session->clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
 	trace_nfs4_destroy_session(session->clp, status);
 
@@ -7347,6 +7343,26 @@ int nfs4_proc_destroy_session(struct nfs4_session *session,
 	return status;
 }
 
+int nfs4_proc_destroy_session(struct nfs4_session *session,
+		struct rpc_cred *cred)
+{
+	unsigned int loop;
+	int ret;
+
+	/* session is still being setup */
+	if (!test_and_clear_bit(NFS4_SESSION_ESTABLISHED, &session->session_state))
+		return 0;
+
+	for (loop = NFS4_MAX_LOOP_ON_RECOVER; loop != 0; loop--) {
+		ret = _nfs4_proc_destroy_session(session, cred);
+		if (ret != -NFS4ERR_DELAY)
+			break;
+		ssleep(1);
+	}
+
+	return ret;
+}
+
 /*
  * Renew the cl_session lease.
  */
-- 
2.3.3


             reply	other threads:[~2015-03-20  8:31 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-20  8:31 Kinglong Mee [this message]
2015-03-22 19:14 ` [PATCH] NFS4: Retry destroy session when getting -NFS4ERR_DELAY Trond Myklebust
2015-03-23  1:16   ` Kinglong Mee
2015-03-23 14:09     ` Trond Myklebust
2015-03-23 16:09       ` Trond Myklebust
2015-03-23 20:21         ` [PATCH v2 1/2] NFSv4: Ensure that we drain the session before shutting it down Trond Myklebust
2015-03-23 20:21           ` [PATCH v2 2/2] NFSv4: Cleanup - move slot_table drain functions into fs/nfs/nfs4session.c Trond Myklebust
2015-03-24 17:00           ` [PATCH v2 1/2] NFSv4: Ensure that we drain the session before shutting it down Kinglong Mee
2015-03-24 17:02             ` Kinglong Mee
2016-08-19 13:41           ` Olga Kornievskaia
2016-08-19 15:45             ` Trond Myklebust

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=550BDAE0.2070409@gmail.com \
    --to=kinglongmee@gmail.com \
    --cc=linux-nfs@vger.kernel.org \
    --cc=trond.myklebust@primarydata.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).