I've deployed openstack using kolla, when I try to launch an instance directly from any image, after some time waiting for Block Device Mapping I get the following error :
I'm confused here, when you say directly from image, do you mean ephemeral volumes (nova) or persistent volumes (cinder)? I will assume cinder volumes since we've BDM here and nova is triggering cinder to create bootable volumes.
Build of instance 4cf01ba2-05b3-44e9-a685-8875d8c96b4e aborted: Volume 01739e82-9e66-41f7-be74-dfbbdcd6746e did not finish being created even after we waited 203 seconds or 61 attempts. And its status is creating.
I've tried increasing block_device_allocate_retries=400 and block_device_allocate_retries_interval=3 , however I keep getting the same error.
But when I create a volume from an image, then launch an instance from that same volume, it works just fine.
Any suggestions for this issue?
Which OpenStack release you're working on?
I think the operation is failing asynchronously on the cinder side (probably in c-vol) and nova times out waiting for a response. I suggest you check cinder logs (c-api, c-sch, c-vol) for a more specific error message.
Thanks and regards
Rajat Dhasmana