https://jira.whamcloud.com This file is an XML representation of an issue en-us 9.4.14 940014 05-12-2023 https://jira.whamcloud.com/browse/LU-9392 Lustre <p>It would be nice if "<tt>lfs setstripe</tt>" and "<tt>lfs migrate</tt>" could pick x out of y given OSTs (y&gt;x). AKA you tell it to stripe across 8 OSTs, give a list of 25 and "lfs migrate" pick the 8 OSTs with most space available. Currently it breaks with x != y.</p> <p>For files with overstriping, "<tt>lfs setstripe -o x,y,z -C N</tt>" should repeat the <tt>x,y,z</tt> pattern until all <tt>N</tt> OSTs are selected.</p> lustre-client-2.7.15.3-3.10.0_229.20.1.el6.x86_64.knl2.x86_64<br/> LU-9392

lfs migrate -o and lfs setstripe -o should pick OST from ost_list

Improvement Minor Open Unresolved WC Triage Michael Hebensteit medium Tue, 4 Apr 2017 01:59:35 +0000 Tue, 1 Aug 2023 23:34:30 +0000 0 6 <p>This would indeed be useful. For permanent collections of "y" OSTs (e.g. different OSS storage types) it is possible to define "OST pools" and then specify a stripe count within the pool.</p> <p>However, being able to use <tt>lfs setstripe -c 3 -o 3,4,13,14,23,24</tt> for temporary collections of OSTs would also be useful, particularly since creating/removing OST pools is limited to the admin.</p> <p>It would also be useful to be able to persistently set <tt>lfs setstripe -o</tt> layouts on a parent directory.</p> <p>It's easy to implement picking stripe_count=x out of y given OSTs (x&lt;=y) by changing the conditions to compare stripe_count with nr_tgts in function lfs_setstripe_internal().</p> <p>What I'm thinking about is the QOS policy. In function lod_alloc_ost_list(), "The exact order of OSTs requested by the user is respected as much as possible depending on OST status." It means, except that the specified OSTs are full or inactive, the OSTs will be selected from the first one in the list strictly. If we want to pick x out of y, do we still want this order? or some disorder?</p> <p>I think in the x &lt; y case it makes sense to use normal RR or QOS to select the starting OST and which other OSTs to use, to preserve space balance. It should be treated in the same way as if the OST list was coming from a pool. </p> <p>The one difficulty that I thought about for the implementation was how to specify both the number of OSTs in the <tt>lov_user_md</tt> and the stripe count for the file? </p> <p>Emoly Liu (emoly@whamcloud.com) uploaded a new patch: <a href="https://review.whamcloud.com/37940" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/37940</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-9392" title="lfs migrate -o and lfs setstripe -o should pick OST from ost_list" class="issue-link" data-issue-key="LU-9392">LU-9392</a> lfs: pick x out of y given OSTs(x &lt;= y) for setstripe<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: ce5835193b6364fc2e853ca140aa502d9e90dddb</p> Related LU-4665 LU-9771 LU-6135 LU-13363 LU-11510 LU-14480 LU-15494 LU-8616 Development Rank 1|hzz97b: Rank (Obsolete) 9223372036854775807