mgr/cephadm: add info to 'ceph orch upgrade status' in cephadm #39880
Conversation
|
NOTE: I noticed when testing this, sometimes right after the mon daemons are upgraded, the container_image_name, container_image_digests and version fields will not be set in the DaemonDescription of some daemons yet when they're checked to see if the daemon needs to be upgraded. In that case the logs will show something like 2021-03-05T21:47:09.196885+0000 mgr.vm-01.cjaxne [INF] Upgrade: Checking mgr daemons even though the mgr daemons have already been upgraded. In that case you could get output like even though the mgr and mon daemons have already been upgraded. I'm not sure if the fields not being set is related to this change. It's not something I've looked for before. I'll look into it more later. |
src/pybind/mgr/cephadm/upgrade.py
Outdated
| @@ -25,6 +25,9 @@ def __init__(self, | |||
| target_version: Optional[str] = None, | |||
| error: Optional[str] = None, | |||
| paused: Optional[bool] = None, | |||
| services_complete: Optional[List[str]] = None, | |||
| progress_str: str = '', | |||
There was a problem hiding this comment.
we should query this from the progress module.
There was a problem hiding this comment.
Would this just give us the % complete being used for the output of 'ceph -s' right now?
There was a problem hiding this comment.
this is already there!
ceph/src/pybind/mgr/cephadm/upgrade.py
Lines 261 to 271 in 543fc00
|
removed all the new fields from the UpgradeStae object. Calculating completed services and number of daemons upgraded when upgrade status is called instead. |
|
can you add |
properly fills in 'services_complete' field, adds info messages to 'message' field such as what daemon type is being upgraded or if we're pulling an image and adds 'progress' field that shows how many daemons have been upgraded so far. Fixes: https://tracker.ceph.com/issues/49235 Signed-off-by: Adam King <adking@redhat.com>
properly fills in 'service_complete' field, adds info messages to 'message' field
such as what daemon type is being upgraded or if we're pulling an image and adds
'progress' field that shows how many daemons have been upgraded so far.
Fixes: https://tracker.ceph.com/issues/49235
Signed-off-by: Adam King adking@redhat.com
"services_complete" field was already present in the output of the command but we never filled it in in cephadm.
"message" field also already existed but was only used for errors. Adding in extra info to say what the upgrade is currently doing. Either what daemon type is being upgraded or that we're pulling an image.
"progress" field is new. It prints out "x/y ceph daemons upgraded" where x is how many daemons we've upgraded so far and y is the total number of daemons we plan to upgrade (daemons whose type is in CEPH_UPGRADE_ORDER). There is already a progress bar the user can see using the "ceph -s" command but I've found it to be inconsistent (https://pastebin.com/raw/JsXkGPkq) so I felt like the orchestrator could use its own.
The idea is to give the user a single command that provides good information on the current state of the upgrade.
Sample output (2 mgr, 3 mon, 3 crash, 6 osd cluster):
Checklist
Show available Jenkins commands
jenkins retest this pleasejenkins test classic perfjenkins test crimson perfjenkins test signedjenkins test make checkjenkins test make check arm64jenkins test submodulesjenkins test dashboardjenkins test apijenkins test docsjenkins render docsjenkins test ceph-volume alljenkins test ceph-volume tox