Synchronize OpenStack objects¶
Fleio is storing a copy of the OpenStack objects in its database. This cache allows Fleio to be faster and puts less load on the OpenStack API.
fleio daemon connects to OpenStack queues and gets push notifications when an OpenStack objects (instances, volumes, networks etc.) are created, modified or deleted.
fleio service processes these notifications and updates the Fleio database.
The synchronization process is running automatically after saving the initial OpenStack configuration and it will populate the Fleio database with OpenStack objects.
If you want to do a manual synchronization, this can be done by running the
After the synchronization is finished, it’s also useful to restart the
systemctl restart fleio
During synchronization Fleio will attempt to make a copy of all supported OpenStack objects into Fleio database. However there are some special cases where objects are treated differently or even ignored.
New in version 2022.05.1.
During various operations (e.g. volume backup) cinder may create temporary volumes that do not count against quota and for whom no notifications are sent. This volumes will not be added to Fleio database during normal operation however if you trigger a sync during a volume backup for example then the temporary volume may be added to the Fleio database until a future sync will remove it.
To prevent this we check consumes_quota attribute for volumes, if consumes_quota is False the volume will be ignored during sync process.
Note that consumes_quota attribute is available only for cinder API version 3.65 or greater, if this API version is not available in your OpenStack deployment we provide a workaround for volume backups (see Exclude backup volumes from sync )