[lustre-discuss] MGTMDT device getting full
Ben Evans
bevans at cray.com
Fri Oct 16 07:01:34 PDT 2015
Looks like you¹ve got some really large changelogs built up. Did you have
robin hood, or some other consumer running at some point that has since
stalled?
-Ben Evans
On 10/16/15, 7:24 AM, "lustre-discuss on behalf of Torsten Harenberg"
<lustre-discuss-bounces at lists.lustre.org on behalf of
torsten.harenberg at cern.ch> wrote:
>Dear all,
>
>I just noticed that the metadata device of our ~120 TB lustre system is
>getting full - see monitoring plot attached.
>
>However, the amount of data stored is more of less the same (usually
>around ~70% of its capacity).
>
>I am trying to undestand what that means. Does a full MGTMDT can be
>translated to "too many files stored"? Or is some kind of clean up needed?
>
>The file system was created with
>
>mkfs.lustre --fsname=lustre --mgs --mdt --backfstype=ext4
>--failnode=132.195.124.201 at tcp --verbose /dev/mapper/MGTMDT
>
>with Lustre 2.1.5.
>
>Recently, the system was updated to 2.5.3
>
>lfs df output is as follows:
>
>[root at fugg1 lustre]# lfs df /lustre
>UUID 1K-blocks Used Available Use% Mounted on
>lustre-MDT0000_UUID 805164976 700243760 51234128 93%
>/lustre[MDT:0]
>lustre-OST0000_UUID 8585168804 5456372044 2699295668 67%
>/lustre[OST:0]
>lustre-OST0001_UUID 8585168804 5243270580 2912396160 64%
>/lustre[OST:1]
>lustre-OST0002_UUID 8585168804 5479924236 2675742656 67%
>/lustre[OST:2]
>lustre-OST0003_UUID 8585168804 5364347704 2791318664 66%
>/lustre[OST:3]
>lustre-OST0004_UUID 8585168804 5420547256 2735119548 66%
>/lustre[OST:4]
>lustre-OST0005_UUID 8585168804 5407878880 2747789400 66%
>/lustre[OST:5]
>lustre-OST0006_UUID 8585168804 5689191524 2466475596 70%
>/lustre[OST:6]
>lustre-OST0007_UUID 8585168804 5541360168 2614306556 68%
>/lustre[OST:7]
>lustre-OST0008_UUID 8585168804 5448642208 2707025228 67%
>/lustre[OST:8]
>lustre-OST0009_UUID 8585168804 5369793176 2785874116 66%
>/lustre[OST:9]
>lustre-OST000a_UUID 8585168804 5461624660 2694043120 67%
>/lustre[OST:10]
>lustre-OST000b_UUID 8585168804 5330093508 2825574244 65%
>/lustre[OST:11]
>lustre-OST000c_UUID 8585168804 5562546324 2593121768 68%
>/lustre[OST:12]
>lustre-OST000d_UUID 8585168804 5559035748 2596632340 68%
>/lustre[OST:13]
>lustre-OST000e_UUID 8585168804 5373777460 2781890244 66%
>/lustre[OST:14]
>
>filesystem summary: 128777532060 81708405476 40626605308 67% /lustre
>
>[root at fugg1 lustre]# lfs df -i /lustre
>UUID Inodes IUsed IFree IUse% Mounted on
>lustre-MDT0000_UUID 536870912 104832073 432038839 20%
>/lustre[MDT:0]
>lustre-OST0000_UUID 16777216 6536109 10241107 39%
>/lustre[OST:0]
>lustre-OST0001_UUID 16777216 6165467 10611749 37%
>/lustre[OST:1]
>lustre-OST0002_UUID 16777216 6709225 10067991 40%
>/lustre[OST:2]
>lustre-OST0003_UUID 16777216 6427098 10350118 38%
>/lustre[OST:3]
>lustre-OST0004_UUID 16777216 6607051 10170165 39%
>/lustre[OST:4]
>lustre-OST0005_UUID 16777216 6725912 10051304 40%
>/lustre[OST:5]
>lustre-OST0006_UUID 16777216 6933544 9843672 41%
>/lustre[OST:6]
>lustre-OST0007_UUID 16777216 6978005 9799211 42%
>/lustre[OST:7]
>lustre-OST0008_UUID 16777216 6252495 10524721 37%
>/lustre[OST:8]
>lustre-OST0009_UUID 16777216 6873340 9903876 41%
>/lustre[OST:9]
>lustre-OST000a_UUID 16777216 7091045 9686171 42%
>/lustre[OST:10]
>lustre-OST000b_UUID 16777216 6337078 10440138 38%
>/lustre[OST:11]
>lustre-OST000c_UUID 16777216 5993369 10783847 36%
>/lustre[OST:12]
>lustre-OST000d_UUID 16777216 6357097 10420119 38%
>/lustre[OST:13]
>lustre-OST000e_UUID 16777216 6087599 10689617 36%
>/lustre[OST:14]
>
>filesystem summary: 536870912 104832073 432038839 20% /lustre
>
>[root at fugg1 lustre]#
>
>Thanks for any hint!
>
>Best regards
>
> Torsten
>
>--
><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
><> <>
><> Dr. Torsten Harenberg Torsten.Harenberg at cern.ch <>
><> Bergische Universitaet <>
><> FB C - Physik Tel.: +49 (0)202 439-3521 <>
><> Gaussstr. 20 Fax : +49 (0)202 439-2811 <>
><> 42097 Wuppertal @CERN: Bat. 1-1-049 <>
><> <>
><><><><><><><>< Of course it runs NetBSD http://www.netbsd.org ><>
More information about the lustre-discuss
mailing list