Try to improve ceph pool creation reliability #12250
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
simondeziel
force-pushed
the
ceph-pool
branch
3 times, most recently
from
99fb30f to
2493696
Compare
Signed-off-by: Simon Deziel <[email protected]>
This is an attempt to improve the reliability of the test suite that often times out during the ceph pool setup: > 2023-09-12T17:14:09.3105279Z + timeout --foreground 120 /home/runner/go/bin/lxc storage create lxdtest-gOl ceph volume.size=25MiB ceph.osd.pg_num=16 --verbose > 2023-09-12T17:16:09.3132429Z + cleanup g_num=16 Signed-off-by: Simon Deziel <[email protected]>
Signed-off-by: Simon Deziel <[email protected]>
`mdl` depends on `core18` and the `microceph` on `core22` Signed-off-by: Simon Deziel <[email protected]>
simondeziel
force-pushed
the
ceph-pool
branch
from
7e22b96 to
3940576
Compare
Signed-off-by: Simon Deziel <[email protected]>
This avoid saving the 404 XML error if the commit hash is wrong/not available: > $ cat gotip.tar.gz > <?xml version='1.0' encoding='UTF-8'?><Error><Code>NoSuchKey</Code><Message> > The specified key does not exist.</Message><Details>No such object: > go-build-snap/go/linux-amd64/<commit-hash>.tar.gz</Details></Error> As that would later on be feed to `tar`/`gzip`: gzip: stdin: not in gzip format tar: Child returned status 1 tar: Error is not recoverable: exiting now Signed-off-by: Simon Deziel <[email protected]>
simondeziel
force-pushed
the
ceph-pool
branch
from
412b025 to
6569581
Compare
tomponline
reviewed
Sep 14, 2023
| sudo microceph disk add --wipe /dev/sdia | ||
| ephemeral_disk="$(findmnt --noheadings --output SOURCE --target /mnt | sed 's/[0-9]\+$//')" | ||
| sudo umount /mnt | ||
| sudo microceph disk add --wipe "${ephemeral_disk}" |
There was a problem hiding this comment.
@simondeziel oh did the microceph folks fix the apparmor issue with apparmor that was preventing the use of partitions?
As sometimes (not always) the github runner ephemeral disk isn't a whole disk but a partition.
There was a problem hiding this comment.
@tomponline yeah, I too noticed the ephemeral disk was not always setup the same way. Sometimes /mnt was a NTFS partition, some other times it was ext4. Sometimes it would be /dev/sdb1, some other /dev/sda1. I've seen NTFS parts at least 2 times and it's likely not good to back a loop-dev. See /dev/sdb1 in blkid's output:
# From https://github.com/canonical/lxd/actions/runs/6177130891/job/16767605386
# but https://github.com/canonical/lxd/actions/runs/6177049637/job/16767324493 also was NTFS
...
+ swapon -s
Filename Type Size Used Priority
/mnt/swapfile file 4194300 0 -2
+ lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
loop0 7:0 0 63.4M 1 loop /snap/core20/1974
loop1 7:1 0 73.9M 1 loop /snap/core22/864
loop2 7:2 0 53.3M 1 loop /snap/snapd/19457
loop3 7:3 0 90.2M 1 loop /snap/microceph/585
sda 8:0 0 86G 0 disk
├─sda1 8:1 0 85.9G 0 part /
├─sda14 8:14 0 4M 0 part
+ blkid
+ sudo swapoff /mnt/swapfile
└─sda15 8:15 0 106M 0 part /boot/efi
sdb 8:16 0 14G 0 disk
└─sdb1 8:17 0 14G 0 part /mnt
/dev/sdb1: LABEL="Temporary Storage" BLOCK_SIZE="512" UUID="24E0FC11E0FBE748" TYPE="ntfs" PARTUUID="f05c5304-01"
/dev/sda15: LABEL_FATBOOT="UEFI" LABEL="UEFI" UUID="D659-5A9F" BLOCK_SIZE="512" TYPE="vfat" PARTUUID="59204cef-8bed-4e09-876e-f620523d8884"
/dev/sda1: LABEL="cloudimg-rootfs" UUID="d4004e88-d528-486d-b994-90c70f9322d4" BLOCK_SIZE="4096" TYPE="ext4" PARTUUID="886ad3da-e4ec-442e-b223-03e615c2b78d"
/dev/loop1: TYPE="squashfs"
/dev/loop2: TYPE="squashfs"
/dev/loop0: TYPE="squashfs"
++ findmnt --noheadings --output SOURCE --target /mnt
++ sed 's/[0-9]\+$//'
+ ephemeral_disk=/dev/sdb
+ sudo umount /mnt
+ sudo microceph disk add --wipe /dev/sdb
As for the part vs whole disk, I didn't think of Apparmor but it could well be. My only thought for Apparmor was the immediate benefit of no longer needing to masquerade a loop device as sdia :) That is moot now that we use /dev/sda or /dev/sdb.
There was a problem hiding this comment.
Lets see if it works, previously microceph's apparmor profile prevented using partitions as disks, but hopefully they fixed that.
There was a problem hiding this comment.
I don't think they fixed it:
+ sudo microceph disk add --wipe /dev/sda1
Error: Failed adding new disk: Failed to wipe the device: Failed to run: dd if=/dev/zero of=/dev/disk/by-id/wwn-0x600224800af2a3230d891ce3c7debbb5-part1 bs=4M count=10 status=none: exit status 1 (dd: failed to open '/dev/disk/by-id/wwn-0x600224800af2a3230d891ce3c7debbb5-part1': Permission denied)
tomponline
approved these changes
Sep 14, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.