[prev in list] [next in list] [prev in thread] [next in thread]
List: lustre-discuss
Subject: Re: [lustre-discuss] MDT mount failing due to unknown param quota
From: Brad Merchant <bmerchant () cambridgecomputer ! com>
Date: 2021-04-03 15:11:02
Message-ID: CAACH-GgjPGXJyeNqc5UJ4D8BmRdw+-R6ttqvWWx_ksARccvRag () mail ! gmail ! com
[Download RAW message or body]
[Attachment #2 (multipart/alternative)]
Thanks Colin, that was it. I'm never quite sure exactly which tunefs
commands need to be run in situations like these, but that sequence worked
perfectly.
Thanks from everyone here for your help!
On Sat, Apr 3, 2021 at 12:32 AM Colin Faber <cfaber@gmail.com> wrote:
> Hi Brad,
>
> Looks like you're suffering from an old bug, you'll need to tuners.lustre
> the MDT target, capture the current params and then --erase-params, --param
> mdd.quota_type=ug and anything else (other than mdt.quota_type=ug) and
> --writeconf
>
> And because of other behavioral issues around --writeconf, you'll need to
> --writeconf (only) all of the OSS targets
>
> And because of behavioural issues with initial mounting of lustre you'll
> want to mount in the specific order of: mgs / MDT, OSS(s) and then clients.
>
> Cheers and good luck.
>
>
> On Fri, Apr 2, 2021, 5:56 PM Brad Merchant <
> bmerchant@cambridgecomputer.com> wrote:
>
>> We have an old IEEL 3.0 system (Lustre ver 2.7.15.3) that we enabled user
>> and group quotas on several months ago. Today we rebooted the MDS node for
>> the first time since then and the MDT refuses to mount with
>>
>>
>> [Fri Apr 2 18:36:33 2021] LDISKFS-fs (dm-1): mounted filesystem with
>> ordered data mode. Opts: user_xattr,errors=remount-ro,no_mbcache
>> [Fri Apr 2 18:36:33 2021] LustreError:
>> 35031:0:(obd_config.c:1345:class_process_proc_param()) mdt: vipbgfs-MDT0000
>> unknown param quota=ug
>> [Fri Apr 2 18:36:33 2021] LustreError:
>> 35031:0:(obd_config.c:1624:class_config_llog_handler()) MGC10.0.0.101@o2ib:
>> cfg command failed: rc = -38
>> [Fri Apr 2 18:36:33 2021] Lustre: cmd=cf00f 0:vipbgfs-MDT0000
>> 1:mdt.quota=ug
>>
>> [Fri Apr 2 18:36:33 2021] LustreError: 15c-8: MGC10.0.0.101@o2ib: The
>> configuration from log 'vipbgfs-MDT0000' failed (-38). This may be the
>> result of communication errors between this node and the MGS, a bad
>> configuration, or other errors. See the syslog for more information.
>> [Fri Apr 2 18:36:33 2021] LustreError:
>> 34940:0:(obd_mount_server.c:1305:server_start_targets()) failed to start
>> server vipbgfs-MDT0000: -38
>> [Fri Apr 2 18:36:33 2021] LustreError:
>> 34940:0:(obd_mount_server.c:1792:server_fill_super()) Unable to start
>> targets: -38
>> [Fri Apr 2 18:36:33 2021] Lustre: Failing over vipbgfs-MDT0000
>> [Fri Apr 2 18:36:33 2021] Lustre: server umount vipbgfs-MDT0000 complete
>> [Fri Apr 2 18:36:33 2021] LustreError:
>> 34940:0:(obd_mount.c:1342:lustre_fill_super()) Unable to mount (-38)
>>
>>
>> Is there a way to disable quota param to get the MDT to mount? Or any
>> other way to get the MDT to mount?
>>
>> If it matters the node is currently running e2fsprogs ver 1.42.13.wc4-7
>> (all targets are ldiskfs)
>>
>>
>> Thanks,
>> Brad Merchant
>> HPC Technologist
>> Cambridge Computer
>> _______________________________________________
>> lustre-discuss mailing list
>> lustre-discuss@lists.lustre.org
>> http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
>>
>
[Attachment #5 (text/html)]
<div dir="ltr">Thanks Colin, that was it. I'm never quite sure exactly which \
tunefs commands need to be run in situations like these, but that sequence worked \
perfectly.<div><br></div><div>Thanks from everyone here for your \
help!</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On \
Sat, Apr 3, 2021 at 12:32 AM Colin Faber <<a \
href="mailto:cfaber@gmail.com">cfaber@gmail.com</a>> wrote:<br></div><blockquote \
class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid \
rgb(204,204,204);padding-left:1ex"><div dir="auto">Hi Brad,<div \
dir="auto"><br></div><div dir="auto">Looks like you're suffering from an old bug, \
you'll need to tuners.lustre the MDT target, capture the current params and then \
--erase-params, --param mdd.quota_type=ug and anything else (other than \
mdt.quota_type=ug) and --writeconf</div><div dir="auto"><br></div><div dir="auto">And \
because of other behavioral issues around --writeconf, you'll need to --writeconf \
(only) all of the OSS targets</div><div dir="auto"><br></div><div dir="auto">And \
because of behavioural issues with initial mounting of lustre you'll want to \
mount in the specific order of: mgs / MDT, OSS(s) and then clients.</div><div \
dir="auto"><br></div><div dir="auto">Cheers and good luck.</div><div \
dir="auto"><br></div></div><br><div class="gmail_quote"><div dir="ltr" \
class="gmail_attr">On Fri, Apr 2, 2021, 5:56 PM Brad Merchant <<a \
href="mailto:bmerchant@cambridgecomputer.com" rel="noreferrer noreferrer" \
target="_blank">bmerchant@cambridgecomputer.com</a>> wrote:<br></div><blockquote \
class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid \
rgb(204,204,204);padding-left:1ex"><div dir="ltr">We have an old IEEL 3.0 system \
(Lustre ver 2.7.15.3) that we enabled user and group quotas on several months ago. \
Today we rebooted the MDS node for the first time since then and the MDT refuses to \
mount with<div><br><div><br></div><div>[Fri Apr 2 18:36:33 2021] LDISKFS-fs (dm-1): \
mounted filesystem with ordered data mode. Opts: \
user_xattr,errors=remount-ro,no_mbcache<br>[Fri Apr 2 18:36:33 2021] LustreError: \
35031:0:(obd_config.c:1345:class_process_proc_param()) mdt: vipbgfs-MDT0000 unknown \
param quota=ug<br>[Fri Apr 2 18:36:33 2021] LustreError: \
35031:0:(obd_config.c:1624:class_config_llog_handler()) MGC10.0.0.101@o2ib: cfg \
command failed: rc = -38<br>[Fri Apr 2 18:36:33 2021] Lustre: cmd=cf00f \
0:vipbgfs-MDT0000 1:mdt.quota=ug <br><br>[Fri Apr 2 18:36:33 2021] LustreError: \
15c-8: MGC10.0.0.101@o2ib: The configuration from log 'vipbgfs-MDT0000' \
failed (-38). This may be the result of communication errors between this node and \
the MGS, a bad configuration, or other errors. See the syslog for more \
information.<br>[Fri Apr 2 18:36:33 2021] LustreError: \
34940:0:(obd_mount_server.c:1305:server_start_targets()) failed to start server \
vipbgfs-MDT0000: -38<br>[Fri Apr 2 18:36:33 2021] LustreError: \
34940:0:(obd_mount_server.c:1792:server_fill_super()) Unable to start targets: \
-38<br>[Fri Apr 2 18:36:33 2021] Lustre: Failing over vipbgfs-MDT0000<br>[Fri Apr \
2 18:36:33 2021] Lustre: server umount vipbgfs-MDT0000 complete<br>[Fri Apr 2 \
18:36:33 2021] LustreError: 34940:0:(obd_mount.c:1342:lustre_fill_super()) Unable to \
mount (-38)<br></div><div><br></div><div><br></div><div>Is there a way to disable \
quota param to get the MDT to mount? Or any other way to get the MDT to \
mount?</div><div><br></div><div>If it matters the node is currently running e2fsprogs \
ver 1.42.13.wc4-7 (all targets are \
ldiskfs)</div><div><br></div><div><br></div><div>Thanks,</div><div>Brad \
Merchant</div><div>HPC Technologist</div><div>Cambridge Computer</div></div></div> \
_______________________________________________<br> lustre-discuss mailing list<br>
<a href="mailto:lustre-discuss@lists.lustre.org" rel="noreferrer noreferrer \
noreferrer" target="_blank">lustre-discuss@lists.lustre.org</a><br> <a \
href="http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org" rel="noreferrer \
noreferrer noreferrer noreferrer" \
target="_blank">http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org</a><br>
</blockquote></div>
</blockquote></div>
_______________________________________________
lustre-discuss mailing list
lustre-discuss@lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic