mgr/cephadm: fix tcmu-runner cephadm_stray_daemon

[melissa-kun-li]

Adds tcmu-runner daemon to managed to prevent cephadm from raising a cephadm_stray_daemon health warning whenever an image is added to a target

Fixes: https://tracker.ceph.com/issues/51111
Signed-off-by: Melissa Limelissali@redhat.com

Checklist

• References tracker ticket
• Updates documentation if necessary
• Includes tests for new functionality or reproducer for bug

---

Show available Jenkins commands

• jenkins retest this please
• jenkins test classic perf
• jenkins test crimson perf
• jenkins test signed
• jenkins test make check
• jenkins test make check arm64
• jenkins test submodules
• jenkins test dashboard
• jenkins test dashboard cephadm
• jenkins test api
• jenkins test docs
• jenkins render docs
• jenkins test ceph-volume all
• jenkins test ceph-volume tox

[melissa-kun-li]

jenkins test make check

[melissa-kun-li]

Test on this branch: melissa-kun-li@30fe259 just with the symlink change, no PR changes:
https://pulpito.ceph.com/melissali-2021-12-01_15:57:33-rbd:iscsi-master-distro-basic-smithi/

2021-12-01T16:28:19.622 INFO:tasks.cram.client.0.smithi026.stdout:/home/ubuntu/cephtest/archive/cram.client.0/gwcli_create.t: failed
2021-12-01T16:28:19.622 INFO:tasks.cram.client.0.smithi026.stdout:--- gwcli_create.t
2021-12-01T16:28:19.623 INFO:tasks.cram.client.0.smithi026.stdout:+++ gwcli_create.t.err
...

2021-12-01T16:28:19.635 INFO:tasks.cram.client.0.smithi026.stdout: Map the LUN
2021-12-01T16:28:19.635 INFO:tasks.cram.client.0.smithi026.stdout: ===========
2021-12-01T16:28:19.635 INFO:tasks.cram.client.0.smithi026.stdout:   $ sudo podman exec $ISCSI_CONTAINER gwcli iscsi-targets/iqn.2003-01.com.redhat.iscsi-gw:ceph-gw/hosts/iqn.1994-05.com.redhat:client disk disk=datapool/block0
2021-12-01T16:28:19.636 INFO:tasks.cram.client.0.smithi026.stdout:+  No such path /iscsi-targets/iqn.2003-01.com.redhat.iscsi-gw:ceph-gw/hosts/iqn.1994-05.com.redhat:client
2021-12-01T16:28:19.636 INFO:tasks.cram.client.0.smithi026.stdout:+  [255]
2021-12-01T16:28:19.636 INFO:tasks.cram.client.0.smithi026.stdout:   $ sudo podman exec $ISCSI_CONTAINER gwcli ls iscsi-targets/ | grep 'o- hosts' | awk -F'[' '{print $2}'
2021-12-01T16:28:19.636 INFO:tasks.cram.client.0.smithi026.stdout:-  Auth: ACL_ENABLED, Hosts: 1]
2021-12-01T16:28:19.636 INFO:tasks.cram.client.0.smithi026.stdout:+  Auth: ACL_ENABLED, Hosts: 0]
2021-12-01T16:28:19.637 INFO:tasks.cram.client.0.smithi026.stdout:   $ sudo podman exec $ISCSI_CONTAINER gwcli ls iscsi-targets/ | grep 'o- iqn.1994-05.com.redhat:client' | awk -F'[' '{print $2}'
2021-12-01T16:28:19.637 INFO:tasks.cram.client.0.smithi026.stdout:-  Auth: None, Disks: 1(300M)]
2021-12-01T16:28:19.637 INFO:tasks.cram.client.0.smithi026.stdout:# Ran 1 tests, 0 skipped, 1 failed.

Test on this branch: melissa-kun-li@2191585 (includes PR changes, deleted rbd/iscsi/supported-random-distros, added symlink to single-container-host.yaml):
https://pulpito.ceph.com/melissali-2021-12-01_06:19:34-rbd:iscsi-master-distro-basic-smithi/6537907/

Same error as the branch without the PR changes.

[ideepika]

@sebastian-philipp @melissa-kun-li ISCSI gateway creation is failing itself somehow

2021-12-01T16:28:19.623 INFO:tasks.cram.client.0.smithi026.stdout:   > IP=`hostname -i | awk '{print $1}'`
2021-12-01T16:28:19.624 INFO:tasks.cram.client.0.smithi026.stdout:   > sudo podman exec $ISCSI_CONTAINER gwcli iscsi-targets/iqn.2003-01.com.redhat.iscsi-gw:ceph-gw/gateways create ip_addresses=$IP gateway_name=$HOST
2021-12-01T16:28:19.624 INFO:tasks.cram.client.0.smithi026.stdout:   $ sudo podman exec $ISCSI_CONTAINER gwcli ls iscsi-targets/ | grep 'o- gateways' | awk -F'[' '{print $2}'

@melissa-kun-li can you run this test on ceph master+symlink change, also would be good to have the symlink change in a different PR.

[melissa-kun-li]

@melissa-kun-li can you run this test on ceph master+symlink change, also would be good to have the symlink change in a different PR.

Ran with just the symlink change on master using my branch new-rbd-iscsi (melissa-kun-li@30fe259) and got the issue in #43833 (comment)

I did:
teuthology-suite -v --ceph-repo https://github.com/ceph/ceph-ci.git --suite-repo https://github.com/melissa-kun-li/ceph.git -c master -m smithi -s rbd/iscsi -k distro -p 50 --suite-branch new-rbd-iscsi -S 0f9ed11e6727bf4fccef4a0481cd9391f27d3cff --force-priority

If there's another thing you meant, please let me know

[ideepika]

@melissa-kun-li can you run this test on ceph master+symlink change, also would be good to have the symlink change in a different PR.

Ran with just the symlink change on master using my branch new-rbd-iscsi (melissa-kun-li@30fe259) and got the issue in #43833 (comment)

I did: teuthology-suite -v --ceph-repo https://github.com/ceph/ceph-ci.git --suite-repo https://github.com/melissa-kun-li/ceph.git -c master -m smithi -s rbd/iscsi -k distro -p 50 --suite-branch new-rbd-iscsi -S 0f9ed11e6727bf4fccef4a0481cd9391f27d3cff --force-priority

If there's another thing you meant, please let me know

Thanks, I will take a look

[ideepika]

@melissa-kun-li I tried reproducing on master, but could not: http://pulpito.front.sepia.ceph.com/ideepika-2021-12-06_19:34:00-rbd:iscsi-wip-yuval-test1-distro-default-smithi/ can you rebase and retry or share the branch you are using for testing?

[melissa-kun-li]

@ideepika thank you for looking into it! I saw your PR: #44243 . I will try testing again. If we can merge this PR, we can look at the rbd/iscsi tests separately

[ideepika]

@ideepika thank you for looking into it! I saw your PR: #44243 . I will try testing again. If we can merge this PR, we can look at the rbd/iscsi tests separately

@melissa-kun-li let me know if you need help with this PR regarding iscsi stuff

[melissa-kun-li]

https://pulpito.ceph.com/melissali-2021-12-25_00:20:55-rbd:iscsi-master-distro-basic-smithi/

[sebastian-philipp]

https://pulpito.ceph.com/swagner-2022-01-06_10:04:35-orch:cephadm-wip-swagner-testing-2022-01-05-1435-distro-default-smithi/

somehow one of the PRs in this run broke NFS and RGW ingress. I really doubt this PR caused this.

[idryomov]

It looks like this introduced a bug:

https://tracker.ceph.com/issues/53830

Note the "No such path /iscsi-targets/iqn.2003-01.com.redhat.iscsi-gw:ceph-gw/hosts/iqn.1994-05.com.redhat:client" error that is mentioned in the previous comments.

Melissa's successful run linked in comment #43833 (comment) didn't actually test this PR's code because it was scheduled against the master branch.

I'll confirm by testing the revert and report back.

[idryomov]

Sorry for the noise, it still fails with this PR taken out. It's just that nothing seems to be have changed on the RBD side in that date range...