[SES6] Simplify management of tuned

[swiftgist]

Disassociate the specific role names from their purpose. Considering that multiple roles can be assigned to a minion, applying a ceph-osd profile to a gateway or having identical profiles for each role is unnecessary.

Signed-off-by: Eric Jackson ejackson@suse.com bnc:1142416

Checklist:

• [ ] Added unittests and or functional tests
• [ ] Adapted documentation
• [ ] Referenced issues or internal bugtracker
• [ ] Ran integration tests successfully (trigger with "@susebot run teuthology" in a GitHub comment; see the wiki for more information)

[swiftgist]

@susebot run teuthology

[susebot]

Commit 3919274bd88a6ae0e5f849df3c3db7556811dd01 is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/338/

[jschmid1]

this implementation causes a bug in stage.3

2019-10-17T14:04:57.814 INFO:teuthology.orchestra.run.target192168000081.stdout:Parsing orchestration ceph.stage.3 steps... fail
2019-10-17T14:04:57.815 INFO:teuthology.orchestra.run.target192168000081.stdout:
2019-10-17T14:04:57.815 INFO:teuthology.orchestra.run.target192168000081.stdout:An error occurred when rendering one of the following states:
2019-10-17T14:04:57.815 INFO:teuthology.orchestra.run.target192168000081.stdout:    - ceph.tuned.throughput (/srv/salt/ceph/tuned/throughput)
2019-10-17T14:04:57.816 INFO:teuthology.orchestra.run.target192168000081.stdout:
2019-10-17T14:04:57.816 INFO:teuthology.orchestra.run.target192168000081.stdout:Error description:
2019-10-17T14:04:57.816 INFO:teuthology.orchestra.run.target192168000081.stdout:  - not ( I@roles:mgr or I@roles:mon ) and I@cluster:ceph: No minions matched the target. No command was sent, no jid was assigned.
2019-10-17T14:04:57.816 INFO:teuthology.orchestra.run.target192168000081.stdout:
2019-10-17T14:04:57.816 INFO:teuthology.orchestra.run.target192168000081.stdout:

[swiftgist]

@susebot run teuthology

[susebot]

Commit 40b14a010705f89da2e325d8387103c0a91edd35 is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/344/

[swiftgist]

@susebot run teuthology

[susebot]

Commit 2aef40a096ff741441a1fc1e4901b7cee68e00f3 is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/345/

[swiftgist]

@susebot run teuthology

[susebot]

Commit ea49e290e47d328f62c8e99978e77b8c4d30752e is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/346/

[jschmid1]

what's the status here @swiftgist

[swiftgist]

Georgios is setting up a teuthology VM.

[swiftgist]

@susebot run teuthology

[susebot]

Commit 5453acd24a0e27a8477ab9a3302548e386846282 is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/350/

[swiftgist]

Previous error was missing a Makefile line... this last one about bluestore seems unrelated

[mgfritch]

Previous error was missing a Makefile line... this last one about bluestore seems unrelated

teuthology.deepsea:tier2.deepsea:tier2/filestore-to-bluestore-migration failures appear fixed by PR #1786

[mgfritch]

@susebot run teuthology

[susebot]

Commit 5453acd24a0e27a8477ab9a3302548e386846282 is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/351/

[swiftgist]

@susebot run teuthology

[swiftgist]

Rebased

[susebot]

Commit 5453acd24a0e27a8477ab9a3302548e386846282 is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/352/

[swiftgist]

@susebot run teuthology

[susebot]

Commit 609fd57e49ad443b80f74b4e8ac166e647a3c08c is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/353/

[swiftgist]

@susebot run teuthology

[susebot]

Commit 609fd57e49ad443b80f74b4e8ac166e647a3c08c is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/356/

[jschmid1]

@swiftgist is this still connected to the tuned changes?

[swiftgist]

The rebuild.node is failing because cephdisks.find_osd_by_id is only checking block and not data. Here's the check:

                    if str(osd_id_search) == str(osd_id) and _lv.tags.get(
                            'ceph.type') == 'block':
                        devs.append(dev)

I thought that was modified at some point, but I don't recall at the moment. The osd.remove call fails when cephdisks.find_by_osd_id returns an empty list.

[jschmid1]

right, this was introduced with c89a914e3 which seems to be not backported to SES6.

edit: from this PR https://github.com/SUSE/DeepSea/pull/1723.

could @gekios please take care of backporting this to SES6?

in the meanwhile you may cherrypick the commit to see if this solves the CI issues @swiftgist

[gekios]

So I opened https://github.com/SUSE/DeepSea/pull/1794 for backporting the commit to SES6 branch.

[jschmid1]

1794 is merged, please rebase @swiftgist

[swiftgist]

rebased

[swiftgist]

@susebot run teuthology

[susebot]

Commit 831b983ff455d647d2695dc8c524f4cf240cf51c is NOT OK for suite deepsea:tier2. Check tests results in the Jenkins job: http://ci.ses.suse.de:8080/job/pr-deepsea/367/

[swiftgist]

Appears that ceph-iscsi failed..... I'm not seeing the connection to tuned

[jschmid1]

This should land in master first