Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merge upstream/release/2.6 into upstream/google/2.6 #15008

Merged
merged 12 commits into from
Aug 27, 2024

Conversation

juszhan1
Copy link
Collaborator

NiuYawei and others added 12 commits August 20, 2024 14:44
…14936)

When a health device is unplugged, we should keep counting the I/O errors
against the device, once the number of I/O errors reached faulty criteria,
the unplugged device should be automatically marked as FAULTY and trigger
targets exclusion accordingly.

Signed-off-by: Niu Yawei <[email protected]>
#14845)

* DAOS-9247 control,bio: Add PCIe link speed and width to NVMe health stats (#14395)

Add NVMe PCIe link speed and width details to the NVMe health stats
returned from dmg storage query list-devices --health and dmg storage
scan --nvme-health commands. The PCIe config space is fetched within
the engine SPDK process (which is functional even when NVMe device is
behind a VMD and bound to a userspace driver) and passed back to the
control-plane over dRPC as a byte-string. The byte-string is formatted
and passed to lspci to convert to human readable text. The output text
is parsed and relevant entries converted into health stat fields to be
printed in dmg output.

Signed-off-by: Tom Nabarro <[email protected]>
--disable-wb-caching should be --disable-wb-cache

Signed-off-by: Jeff Olivier <[email protected]>
…14534) (#14666)

Python buffered read/write relies on st_blksize as default buffer size. dfuse calls dfs_ostatx() which returns stat with zero st_blksize. Consequently, 4096 for st_blksize is returned to application.

Signed-off-by: Lei Huang <[email protected]>
support dynamic stonewall file for ior and mdtest by using
DAOS_TEST_LOG_DIR

Signed-off-by: Dalton Bohning <[email protected]>
Update il -> ioil for consistency

Signed-off-by: Dalton Bohning <[email protected]>
…#14665) (#14966)

In the event that NVMe PCIe link negotiated speed or width does not
match the maximum capability of the device, a RAS event will be raised
(and published in syslog) during a health stat scan (dmg storage query
list-devices --health or dmg storage scan --nvme-health). Further
events will be raised on link state changes. With this change, RAS
events are only raised during a health stat scan. Future changes may
be made to raise RAS events in real time for PCIe link state changes.

Signed-off-by: Tom Nabarro <[email protected]>
- Fix finalization order in the cart_ctl tool, where now
a group is destroyed only after the progress has been stopped.

This avoids a segfault that can happen if an RPC is cancelled
as part of the 'progress stop' sequence, which then references
a group.

Signed-off-by: Alexander A Oganezov <[email protected]>
#14960)

Create network context when a read/write operation is encountered.

Signed-off-by: Lei Huang <[email protected]>
…/2.6

Required-githooks: true

Change-Id: Ifce37d0757049937230e1e2a564b1b55bb2a3110
Copy link

Errors are component not formatted correctly,Ticket number prefix incorrect,PR title is malformatted. See https://daosio.atlassian.net/wiki/spaces/DC/pages/11133911069/Commit+Comments,Unable to load ticket data
https://daosio.atlassian.net/browse/Merge

@juszhan1 juszhan1 requested review from jolivier23 and mjmac August 26, 2024 16:58
@juszhan1 juszhan1 merged commit 18358ab into google/2.6 Aug 27, 2024
68 of 70 checks passed
@juszhan1 juszhan1 deleted the juszhan/google/2.6 branch August 27, 2024 01:20
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Development

Successfully merging this pull request may close these issues.

9 participants