From: Marc Eshel <eshel@almaden.ibm.com>
To: Benny Halevy <bhalevy@panasas.com>
Cc: Boaz Harrosh <bharrosh@panasas.com>,
Tigran Mkrtchyan <tigran.mkrtchyan@desy.de>,
NFS list <linux-nfs@vger.kernel.org>,
linux-nfs-owner@vger.kernel.org, Fred Isaman <iisaman@netapp.com>
Subject: Re: pNFS DS session
Date: Fri, 01 Oct 2010 08:40:36 -0700 [thread overview]
Message-ID: <4CA600F4.2010006@almaden.ibm.com> (raw)
In-Reply-To: <4CA5D537.30300@panasas.com>
On 10/1/2010 5:33 AM, Benny Halevy wrote:
> On 2010-10-01 10:47, Boaz Harrosh wrote:
>> On 10/01/2010 08:12 AM, Tigran Mkrtchyan wrote:
>>> On 10/01/2010 06:17 AM, Marc Eshel wrote:
>>>> Hi Benny,
>>>>
>>>> Running connectathon I see that some times the clients decides to destroy
>>>> the session with the DS. The test continue and the session is
>>>> re-established. It looks like layout return reduces the hold on device
>>>> info the reduces the hold on the client struct which decide to destroy the
>>>> session. Is that a known problem?
>>>>
>> Yes, I want to emphasize on Marks words: "a known *problem*"
> Marc, assuming the code behaves as expected, does this cause any other badness
> like the GETATTRs you see going out to the DS?
>
> Benny
>
No i don't see any "badness" the test continues without errors and this
problem is not related to the GETATTRs I see on the DS but I would
consider destroying the session in short run of couple of minutes some
times more than one time as something bad.
Marc.
>> I have objected strongly to this new "fixture" by Fred and backed up by Benny.
>> They decided to only hold a deviceid as long as a layout references it. And
>> not like before, until unmount. I think they are totally wrong with regard
>> to current servers and implementation and a setup that might enjoy what is
>> done now, will not exist for at least 3 years or more.
>>
>> An argument was made that current code is more simple. But I have demonstrated
>> that all that is needed is a one-line get_ref at device add. And that iteration
>> on all devices at umount time that was there before and removed in latest code.
>>
>> !!!
>> Boaz
>>
>>
>>> Just wan to confirm that we see this as well (pnfs-submit and
>>> pnfs-all-latest).
>>>
>>> Tigran.
>>>> Thanks, Marc.
>>>> --
>>>> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
>>>> the body of a message to majordomo@vger.kernel.org
>>>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>>>
>
next prev parent reply other threads:[~2010-10-01 15:40 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-08-30 16:34 pnfs git tree status pnfs-all-2.6.36-rc3-2010-08-30 Benny Halevy
2010-08-30 16:43 ` [PATCH 1/2] SQUASHME: pnfs-submit: do not use NIPQUAD in nfs4_callback_layoutrecall dprintk Benny Halevy
2010-08-30 16:43 ` [PATCH 2/2] SQUASHME: pnfs-submit: get machince creds for getdeviceinfo Benny Halevy
2010-08-30 16:44 ` [PATCH 3/3] SQUASHME: pnfs: get machince creds for getdevicelist Benny Halevy
2010-08-30 16:44 ` [PATCH 4/4] SQUASHME: pnfs-obj use REQ flags rather than BIO flags Benny Halevy
2010-09-29 11:06 ` pnfs git tree status pnfs-all-2.6.36-rc6-2010-09-29 Benny Halevy
2010-09-29 11:09 ` [PATCH 1/5] SQUASHME: pnfs-submit: add missing include file in nfs4filelayoutdev.c Benny Halevy
2010-09-29 11:09 ` [PATCH 2/5] SQUASHME: move nfs4_deviceid definitions to include/linux/nfs4.h Benny Halevy
2010-09-29 11:09 ` [PATCH 3/5] SQUASHME: pnfs-submit: handle non-pnfs case in set_pnfs_layoutdriver Benny Halevy
2010-09-29 11:10 ` [PATCH 4/5] pnfs-submit: file needs layout commit, server attributes may be stale Benny Halevy
2010-09-29 11:10 ` [PATCH 5/5] NFS: clear fsinfo before sendign rpc Benny Halevy
2010-10-13 18:03 ` Fred Isaman
2010-10-13 20:34 ` Benny Halevy
2010-10-13 20:54 ` Trond Myklebust
2010-10-13 21:13 ` Benny Halevy
2010-10-13 21:20 ` Trond Myklebust
[not found] ` <1287004813.3015.212.camel-rJ7iovZKK19ZJLDQqaL3InhyD016LWXt@public.gmane.org>
2010-10-13 21:27 ` Benny Halevy
2010-09-29 11:10 ` [PATCH 6/9] pnfs: alloc and free layout_hdr layoutdriver methods Benny Halevy
2010-09-29 11:10 ` [PATCH 7/9] pnfs: allow nfs4_proc_layoutget to sleep on invalid lsegs Benny Halevy
2010-09-29 11:10 ` [PATCH 8/9] SQUASHME: pnfs: refactor put_lseg{_locked} Benny Halevy
2010-09-29 11:11 ` [PATCH 9/9] SQUASHME: pnfs: get rid of lo_waitq Benny Halevy
2010-09-29 11:11 ` [PATCH 10/15] SQUASHME: pnfsblock: remove obsolete include file from blocklayout.h Benny Halevy
2010-09-29 11:11 ` [PATCH 11/15] SQUASHME: pnfsblock: use nfs4_deviceid Benny Halevy
2010-09-29 11:11 ` [PATCH 12/15] SQUASHME: pnfsblock: no callback ops Benny Halevy
2010-09-29 11:11 ` [PATCH 13/15] SQAUSHME: pnfsblock: no PNFS_NFS_SERVER Benny Halevy
2010-09-29 11:12 ` [PATCH 14/15] SQUASHME: pnfsblock: no dev_notify_types Benny Halevy
2010-09-29 11:12 ` [PATCH 15/15] SQUASHME: pnfsblock: use new struct pnfs_layout_hdr Benny Halevy
2010-09-29 11:12 ` [PATCH 16/17] SQUASHME: pnfs-obj: fix REQ flags usage Benny Halevy
2010-09-29 11:12 ` [PATCH 17/17] SQUASHME: pnfs-obj: convert to new pnfs-submit changes Benny Halevy
2010-09-29 11:12 ` [PATCH 18/18] SQUASHME: pnfsd-exofs: " Benny Halevy
2010-09-29 11:17 ` pnfs git tree status pnfs-all-2.6.36-rc6-2010-09-29 Benny Halevy
2010-09-29 11:18 ` [PATCH] pnfs: get layout in proper segments Benny Halevy
2010-09-29 14:07 ` pnfs git tree status pnfs-all-2.6.36-rc6-2010-09-29 Tigran Mkrtchyan
2010-09-29 14:09 ` Tigran Mkrtchyan
2010-09-29 14:22 ` Benny Halevy
2010-09-29 14:36 ` Tigran Mkrtchyan
2010-09-29 15:25 ` Benny Halevy
2010-09-29 16:57 ` Tigran Mkrtchyan
2010-09-29 18:35 ` Fred Isaman
2010-09-30 7:58 ` Tigran Mkrtchyan
2010-09-30 8:29 ` Benny Halevy
2010-09-30 8:30 ` Benny Halevy
2010-09-30 9:12 ` Tigran Mkrtchyan
2010-09-30 9:17 ` Benny Halevy
2010-09-30 21:37 ` pnfs-all-2.6.36-rc6-2010-09-30 Benny Halevy
2010-09-30 22:23 ` pnfs git tree status pnfs-all-2.6.36-rc6-2010-09-29 Marc Eshel
2010-10-01 4:17 ` pNFS DS session Marc Eshel
2010-10-01 6:12 ` Tigran Mkrtchyan
2010-10-01 8:47 ` Boaz Harrosh
2010-10-01 12:33 ` Benny Halevy
2010-10-01 15:40 ` Marc Eshel [this message]
2010-10-01 17:10 ` J. Bruce Fields
2010-10-01 18:00 ` Marc Eshel
2010-10-01 18:14 ` J. Bruce Fields
2010-10-01 18:29 ` Andy Adamson
2010-10-03 14:55 ` Boaz Harrosh
[not found] ` <4CA3C1E2.2050701@RedHat.com>
2010-09-30 8:26 ` pnfs git tree status pnfs-all-2.6.36-rc6-2010-09-29 Benny Halevy
2010-10-28 13:36 ` pnfs git tree status pnfs-all-2.6.36-2010-10-28 Benny Halevy
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=4CA600F4.2010006@almaden.ibm.com \
--to=eshel@almaden.ibm.com \
--cc=bhalevy@panasas.com \
--cc=bharrosh@panasas.com \
--cc=iisaman@netapp.com \
--cc=linux-nfs-owner@vger.kernel.org \
--cc=linux-nfs@vger.kernel.org \
--cc=tigran.mkrtchyan@desy.de \
/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).