Good morning / Good evening people! I'm working wi...
# citrix-cloud
a
Good morning / Good evening people! I'm working with a new non-persistent multi-session environment in Citrix Cloud where the only thing onprem is the VDAs. Configured Citrix Cloud MCS and Citrix Cloud WEM together with profile containers (also configured in Citrix cloud, tried both through policies and WEM). Whenever I login, I feel like the WEM sync is very unstable. What do I mean by "sync"? Lets say I login with my user, a profile container is created and the start menu which is coming from WEM is not always syncing. Sometimes it does however and it works fine. When it works, I simulate a user cleanup or removal by the support and I do that by deleting the user folder with the UPM container. Once that is done and I login again, I get a new container. However, I see weird issues like search bar not being responsive at all or the start menu not being fetched from WEM. When this happens, if I logout and login again, everything is synced and works and looks good again. Its as if WEM doesn't apply when the profile is created for the first time but it applys on re-login. This doesn't feel stable at all, it feels like if I let 100 users login, some of them will have the correct start menu and policies, others not. Some will have a broken search bar, others not. I can't ask them to re-login when they have these issues, doesn't feel professional. Do you guys use a similar setup where it works well for you? I just don't understand why this is so unstable with Citrix Cloud. I have put the WEM database and agent files on the MCS cache disk which is created by MCS in Citrix Cloud. Confirmed that they are there, WEM is configured with both OU & Machine catalog (configuration set). Everything seems to be fine but the environment feels very unstable. I can't let this go live, it needs to be way more responsive. I know and have seen that WEM onprem is very stable and responsive. One thing I have noted is that I have two DBs, one locally under C and another on the persistent E disk which is created by MCS which Im pointing to through the norskale registry key. Not sure why it is still creating a DB under C when I'm pointing to E: - on my master image I have removed the db folder under C and pointing to E: in the registry. However I have no E partition because MCS in Citrix cloud is creating the E partition where it puts the MCS cache.
image.png
I think I might have found the issue... Citrix Cloud MCS is changing letter on the partitions randomly even though I clearly configured the MCS cache as E: - need a way to make sure that the partition letters are persistent because right now they are changing wildly on reboot, both the WEM cache partition letter and the MCS cache partition letter
I have now separate partitions for MCS cache and WEM cache but when the VDA starts with the new image, it will still put the wem cache on the MCS partition E even though I have configured W under norskale in registry
makes no sense.. there is nothing that points to E, no GPO nothing. Why does the cache end up there?? if I restart the WEM service, it will appear under W. So now I have cache under E, W and C aswell
I ended up removing the WEM cache partition so now there is only 1 partition. However, MCS is still failing and giving its partition the letter D even though I have configured E in Citrix Cloud. This is pure bug. Now that I have no WEM cache, nothing configured as alternative cache location in registry, the WEM cache is created under C where it usually resides per default. However, even here I see that whenever I login for the first time the search bar gets bugged(not able to click on it or search) and then if I logoff, login again everything works perfectly fine. So this issue appears when removing the profile folder and then I login again, get a new container and profile but start menu is bugged and I have to re-login to fix it. Very weird..
j
How often is the WEM settings modified? If not frequently, then just include the WEM cache on the C: drive and make sure your maintenance machine is part of the same configuration set. Then in your sealing process of the maintenance image run the agentcacheutility to update the local database.