Business Continuity with FSLogix O365 Container - HA and DR Patterns and Considerations

[shocko]

We are using FSLogix O365 Container (version 2.9.8361.52326) on Citrix Virtual Apps and desktops LTSR 2203 CU2 provisioned via Citrix PVS 2203 CU2. The VHDX files are mounted over CIFS from a Netapp C-Mode Ontap 9.3 cluster. We recently had an Ontap failure resulting it what seems to have been Outlook OST corruption (but not VHDX corruption). It got me thinking around backup/restore for the FSLogix solution.

 

What options are available to us? Since the VHDX files are mounted from the CIFS share any network interruption could potentially corrupt them I would think. Taking backups at the storage layer via Netapp WAFL snapshots also seems a poor choice as no guarantee the VHDX files therein are consistent at backup time as they could be mounted to a host  and being written to at that point in time. 

In our case we ended up having to try and automate detection of the corrupt OST file and gave up on that quickly and simply removed the users VHDX file. This worked but on next logon their OST populated in Outlook cached mode and this took a long time (45 mins +) for each user as they mostly have large mailboxes (30 GB+)

I wonder what others are doing in this space for HA/DR around FSLogix O365 containers? 

[James Kindon]

your simplest option to protect against short term network outages is Cloud Cache with FSLogix, Citrix Containers have some file based resiliency in front of their containers in the mix too but I haven't tested that with OST