Azure Service Fabric deployments consume a lot disk space -
i operate on-premise azure service fabric cluster testing purposes. consists of 3 nodes, running on single virtual machine (windows server 2012) 50 gb disk attached it.
further set continuous deployment tfs release pipeline cluster. after approx. 80 deployments, service fabric consumed available disk space , further deployments fail.
most of space taken c:\programdata\sf\data, took around 28gb, while each code package has size of ~130 mb. after have unprovisioned many of old deployments (manually via sf portal), around 5gb released. many of old files still around in c:\programdata\sf\data.
what best approach improve this?
- why files old deployments still on disk after unprovisioning?
- is possible delete these files manually?
- is possible automate deprovisioning?
on production environment situation should relaxed anyhow (since there 1 node per machine , bigger disks). nevertheless put off evil day. feel safer avoid situation @ all.
edit
it seems sf deleting deployment packages delay. checked test cluster after 1 day, , unprovisioned packages vanished finally.
it seems sf deleting deployment packages delay. checked test cluster after 1 day, , unprovisioned packages vanished finally.
further found unregister-servicefabricapplicationtype cmdlet automate unprovisioning process (https://msdn.microsoft.com/en-us/library/mt125885.aspx).
Comments
Post a Comment