[lustre-discuss] Quota values showing as inaccurate

[Scott Wood]

Thanks Andreas,

That highlights the cause.  It looks like two OSTs were not cleanly deactivated. OST0048 and OST0049 should be gone.  I must have missed them.  My mistake!   I'll re-run

lctl conf_param qimrb-OST0048.osc.active=0
lctl conf_param qimrb-OST0049.osc.active=0

Should that be done during an outage?  After fixing that up, will I need to quota_slave.force_reint=1 all the OSTs or, based on the fact that only 48 and 49 have [0], should I be good

For reference, lfs quota -v follows.

Cheers!
Scott

[root at lustreclient ~]# lfs quota  -v -u somedude /mnt/lustre

Disk quotas for usr somedude (uid 10012):
     Filesystem  kbytes   quota   limit   grace   files   quota   limit   grace
    /mnt/lustre [5024896]       0       0       - 6812714       0       0       -
lustre-MDT0000_UUID
                5024896       -       0       - 6812714       -       0       -
quotactl ost0 failed.
quotactl ost1 failed.
...
many more deactivated OSTs
...
lustre-OST0048_UUID[inact]
                    [0]       -     [0]       -       -       -       -       -
lustre-OST0049_UUID[inact]
                    [0]       -     [0]       -       -       -       -       -
...
many more deactivated OSTs
...
lustre-OST0084_UUID
                250701456       -       0       -       -       -       -       -
...
all the rest of the active OSTs
...
lustre-OST0103_UUID
                1527490460       -       0       -       -       -       -       -
Total allocated inode limit: 0, total allocated block limit: 0
uid 10012 is using default block quota setting
uid 10012 is using default file quota setting
Some errors happened when getting quota info. Some devices may be not working or deactivated. The data in "[]" is inaccurate.
________________________________
From: Andreas Dilger <adilger at thelustrecollective.com>
Sent: Friday, 28 November 2025 4:03 PM
To: Scott Wood <woodystrash at hotmail.com>
Cc: lustre-discuss <lustre-discuss at lists.lustre.org>
Subject: Re: [lustre-discuss] Quota values showing as inaccurate

Scott,
you should check "lfs quota -v -u XXXXX" to see to which OSTs the quota is assigned. Some might be assigned to the old OSTs?

Cheers, Andreas

On Nov 27, 2025, at 22:15, Scott Wood via lustre-discuss <lustre-discuss at lists.lustre.org> wrote:


Hi All,

We've recently drained and removed 72 OSTs as per 14.9.3. Removing an OST from the File System.  Since doing so, our quotas are showing as inaccurate.  We don't enforce, but we do use them for monitoring use so we'd like to get them happy again.  We have a scheduled maintenance window coming up and I'm curious if there are any steps we should take during the downtime to help the system reconcile the quota values.

We're ldiskfs on lustre 2.15.5.  Depending on how busy the outage is, we may upgrade to 2.15.7 but we may not be able to fit it in.

Example:

[root at lustreclient ~]# lfs quota -u somedude /mnt/lustre/
Disk quotas for usr somedude (uid 10012):
     Filesystem  kbytes   quota   limit   grace   files   quota   limit   grace
   /mnt/lustre/ [5024896]       0       0       - 6812714       0       0       -
uid 10012 is using default block quota setting
uid 10012 is using default file quota setting
Some errors happened when getting quota info. Some devices may be not working or deactivated. The data in "[]" is inaccurate.

Cheers
Scott
_______________________________________________
lustre-discuss mailing list
lustre-discuss at lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20251128/9a8fca84/attachment-0001.htm>