Openstack - Flavor <flavor id> could not be found #12440
Comments
|
I am also having save issue, Could any one please help |
|
This should be fixed as of: ManageIQ/manageiq-providers-openstack#1 - are you using the latest code? |
|
Hello tzumainn, later that I tried with cloudform4.5 (cfme-appliance-5.8.0.17-1.el7cf.x86_64) and there the newly recommended changed already persist. evm.log [----] E, [2017-06-06T18:15:24.489266 #14194:da3130] ERROR -- : MIQ(ManageIQ::Providers::Openstack::CloudManager::Refresher#refresh) EMS: [RDU2], id: [58000000000007] Refresh failed Note: the flavor is not there and not attched to any instaces. Your help really appraciate |
|
That's really odd! The failure happens when trying to reach this URL: /v2.1/f3262680df7349c988459a950c18a1d0/flavors/96f1a662-b440-4b60-bf94-ffab1d83a649/os-flavor-access.json Which looks like it's looking for something flavor-related in relation to a cloud object with ID f3262680df7349c988459a950c18a1d0 Is there a way to identify what f3262680df7349c988459a950c18a1d0 might be - looking through instance lists, project lists, etc? |
|
We have admin project with id |
|
Ah - in retrospect I should have looked at the stack trace a bit more carefully, as it points to the line with the exact issue: Short-term fix would probably be to delete the flavor from the ManageIQ database; that should get things working right away. We'll work on a permanent bug fix. |
|
@psachin Sorry, one more question - can you check if a flavor with that ID exists in the database? |
|
@tzumainn flavor with id not present in the database, its shows deleted. MariaDB [nova]> select name,deleted,flavorid,name,deleted_at,updated_at,created_at from instance_types where flavorid='96f1a662-b440-4b60-bf94-ffab1d83a649'; MariaDB [nova]> Is the bug/RFE opened for same or do you want me to open new RFE? |
|
Ah, I see. So technically the flavor is present (and the right sort of OpenStack API call can return it); but it's essentially not-present for any tenant, leading to this error. The solution would probably be to filter out flavors wit the deleted flag. Could you open up a bug report for this? |
|
I've not been able to reproduce this error on Euwe, Fine, or the latest master. Any other information you can provide about your environment to help me reproduce the problem would be a great help. |
|
@mansam In my case, OpenStack refresh failed in initial refresh cycle with the above-mentioned image. As I said that flavor was deleted from osp and also not available in db. Note: there are many flavors deleted from osp but having issue with some only like id 96f1a662-b440-4b60-bf94-ffab1d83a649
|
|
Hey guys, I'm seeing the same issue for Euwe-1. I've hit the "Flavour could not be found" with a 404, which is breaking the OSP provider refresh (Keystone v3, with domains). I created a flavour, provisioned an instance from the flavour, and deleted the flavour. I also deleted the flavour from the ManageIQ DB, and still seeing the error. Any ideas? |
flavor
My initial guess was deletion flag was boolean. But I agree, any flavor which is marked with non-zero is deleted flavor. |
|
@deepakhajare Did I understand correctly from your above post that you had the problem even with cfme-appliance-5.8.0.17-1.el7cf.x86_64? |
|
Yes it's true we face this with
appliance-5.8.0.17-1.el7cf.x86_64.
And the older release .
—
On 07-Jun-2017 10:55 PM, "Samuel Lucidi" <notifications@github.com> wrote:
@deepakhajare <https://github.com/deepakhajare> Did I understand correctly
from your above post that you had the problem even with
cfme-appliance-5.8.0.17-1.el7cf.x86_64?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#12440 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/ADDkYu7vnWouwziL4a1GQqshH6Q9uSQXks5sBtz3gaJpZM4Kpphm>
.
|
|
When you tried with appliance-5.8.0.17-1.el7cf.x86_64, was it a fresh database? |
|
I work with @deepakhajare. Yes it was fresh database. |
|
We're having difficulties reproducing this error in our environments. Would it be possible to try a direct change to the code in your environment? Specifically, to catch an error on this line: And just return [] if there's an error? |
|
I'd just add that you will also want to include L49 in the same rescue block you use to catch the error on L48 while testing this. |
|
Replacing lines 48-49 with worked for me. |
|
Thanks for the feedback, @vivobg. Hi @deepakhajare and |
|
@cankarm is this still a valid issue. If not can you close. |
|
I haven't tried this for along time. I'm closing. Thanks for noticing @JPrause . |
Problem: Refresh Provider does not work after the flavor in Openstack is removed/changed.
The appliance release affected: http://releases.manageiq.org/manageiq-openstack-euwe-1-rc1.qc2
Description: If you change the flavour in Openstack (i.e. accessibility,etc), the ID of the flavor changes. ManageIQ is still expecting to find the old non-existant flavor. Here is the error code from ems.log:
The text was updated successfully, but these errors were encountered: