[LU-5216] HSM: restore doesn't restart for new copytool

• Type: Bug
• Resolution: Unresolved
• Priority: Major
• Affects Version/s: Lustre 2.5.1
• Labels: patch
• Environment: Centos 6.5, Lustre 2.5.56
• Severity: 3
• Assignee: Jean-Baptiste Riaux (Inactive)
• Reporter: Frank Zago (Inactive)

Description

When a restoration doesn't complete properly by a copytool, it is not restarted, and processes get stuck in Lustre modules.

For instance, create a large file that takes a few seconds to archive/restore, then release it, and access it:

# dd if=/dev/urandom of=/mnt/lustre/bigf bs=1M count 1000
# lfs hsm_archive /mnt/lustre/bigf
# lfs hsm_release  /mnt/lustre/bigf
# sleep 5https://jira.hpdd.intel.com/browse/LU-5216#
# md5sum /mnt/lustre/bigf

During the restoration, kill the copytool, so no complete event is sent to the MDS.

Note that at this point, it is possible the copytool is unkillable, and the only fix is to reboot the client running that copytool.

When the copytool restarts, nothing happens. The process trying to read the file is stuck (apparently forever) there:

# cat /proc/1675/stack 
[<ffffffffa09dc04c>] ll_layout_refresh+0x25c/0xfe0 [lustre]
[<ffffffffa0a28240>] vvp_io_init+0x340/0x490 [lustre]
[<ffffffffa04dab68>] cl_io_init0+0x98/0x160 [obdclass]
[<ffffffffa04dd794>] cl_io_init+0x64/0xe0 [obdclass]
[<ffffffffa04debfd>] cl_io_rw_init+0x8d/0x200 [obdclass]
[<ffffffffa09cbe38>] ll_file_io_generic+0x208/0x710 [lustre]
[<ffffffffa09ccf8f>] ll_file_aio_read+0x13f/0x2c0 [lustre]
[<ffffffffa09cd27c>] ll_file_read+0x16c/0x2a0 [lustre]
[<ffffffff81189365>] vfs_read+0xb5/0x1a0
[<ffffffff811894a1>] sys_read+0x51/0x90
[<ffffffff8100b072>] system_call_fastpath+0x16/0x1b
[<ffffffffffffffff>] 0xffffffffffffffff

It is also unkillable.

If I issue the "lfs hsm_restore bigf.bin", lfs gets stuck too:

# cat /proc/1723/stack 
[<ffffffffa06ad29a>] ptlrpc_set_wait+0x2da/0x860 [ptlrpc]
[<ffffffffa06ad8a7>] ptlrpc_queue_wait+0x87/0x220 [ptlrpc]
[<ffffffffa0868913>] mdc_iocontrol+0x2113/0x27f0 [mdc]
[<ffffffffa0af5265>] obd_iocontrol+0xe5/0x360 [lmv]
[<ffffffffa0b0c145>] lmv_iocontrol+0x1c85/0x2b10 [lmv]
[<ffffffffa09bb235>] obd_iocontrol+0xe5/0x360 [lustre]
[<ffffffffa09c64d7>] ll_dir_ioctl+0x4237/0x5dc0 [lustre]
[<ffffffff8119d802>] vfs_ioctl+0x22/0xa0
[<ffffffff8119d9a4>] do_vfs_ioctl+0x84/0x580
[<ffffffff8119df21>] sys_ioctl+0x81/0xa0
[<ffffffff8100b072>] system_call_fastpath+0x16/0x1b
[<ffffffffffffffff>] 0xffffffffffffffff

However lfs hsm operation still works on other files.

Issue Links

• is blocking: LU-10175 DoM:Full support for the LDLM lock convert
• is related to: LU-11284 Full lock convert conflicts with HSM
• is related to: LU-8905 tests: sanity-hsm test_3[3-6] does not use ps correctly

Comments

[Ben Evans (Inactive)]

Sergey, I've got a patch worked up for the second part of the patch, but it depends on the first.  As soon as the first lands, I'll push the second one.

[Sergey Cheremencev]

Could someone clarify the question about the second part of patch ?
Do you have plans for land it ?

Right now the 1nd part of patch (https://review.whamcloud.com/#/c/31105/) doesn't solve the issue with killing CopyTool on agent. Original sanity-hsm_62 that demonstrated the problem is changed beginning from patchset 3. And it is the reason why it passes - before patchset 3 it always failed.

If right after the 1st part of the patch you have plans to land the 2nd I can prepare and send this patch. Suggest to add 2 sanit-hsm_62 tests - "a - Evicting a client should cancel its requests" and "b - Stopping a copytool should cancel its requests" for the 2nd part of the patch.

Please let me know your position.

Thanks Sergey

[Cory Spitz]

We discussed this at the 3/8 LWG. JB, we'll try to help you land your patch and work out any issues with John H. Thanks.

[Peter Jones]

spitzcor why don't we discuss at the upcoming LWG meeting?

[Cory Spitz]

pjones, I'd like to get this on the 2.11.0 radar. May we set the Fix Version? bevans & sergey from Cray will follow with more to better kick-start the conversation.

[Jean-Baptiste Riaux (Inactive)]

New patch (https://review.whamcloud.com/#/c/31105/) available including only first part of previous patch (https://review.whamcloud.com/24238/ landed and reverted https://review.whamcloud.com/30615/).

This patch keeps only the first fix:
Unexpected client (data mover node) eviction could
cause on going hsm requests to be stuck in "STARTED"
state as the copy tool running on the data mover node
is not available anymore and requests could not be
finished. This patch unregisters the copy tool and
cancels all the requests on the copytool's agent.

[Gerrit Updater]

Jean-Baptiste Riaux ([e-mail]) uploaded a new patch: https://review.whamcloud.com/31105
Subject: LU-5216 hsm: cancel hsm actions running on CT when killed
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: c3d17bd9e25d5d27ba55835801798e2fdea36c49

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/30615/
Subject: Revert "LU-5216 hsm: cancel hsm actions running on CT when killed"
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: be4507fb45074ad24208c494f98a00da90b13665

[Gerrit Updater]

John L. Hammond ([e-mail]) uploaded a new patch: https://review.whamcloud.com/30615
Subject: Revert "LU-5216 hsm: cancel hsm actions running on CT when killed"
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: a57dcdbca18f9721169f06c312519c619856a656

[Peter Jones]

Landed for 2.11