Last updated:
Recently, with WEM 2507 CU1 on non-persistent VDAs, I encountered behavior that felt very familiar.
And yes, I observed it myself.
The VDA starts correctly, registers and initially appears completely healthy.
But the WEM Agent cache does not initialize correctly, so the WEM configuration is not applied reliably.
Interestingly, similar cases are currently appearing in the community:
Golden image → WEM works ✅
Provisioned VDA → WEM cache issue ❌
Cache reinitialize → WEM works again ✅
According to a community report, Citrix Support also suspected a startup timing issue: the VDA and WEM Agent start before the local WEM cache database has initialized correctly.
The interesting part:
This topic is not new at all.
Back in 2017, BIS-F introduced a WEM cache refresh at system startup.
A feature added later is particularly interesting for non-persistent images:
“Move WEM Agent Cache to the PVS or MCSIO persistent disk”
This places the WEM cache on a persistent location rather than the non-persistent system disk, for example:
PVS/MCSIO CacheDisk
↓
Citrix
↓
WEMCache
At system startup, BIS-F can also run commands such as:
-RefreshCache -RefreshSettings -RefreshCache -Reinitialize
.
To be honest, I had not paid much attention to this functionality for a long time, until I observed the behavior again myself.
A useful reminder:
Just because an old workaround once appeared unnecessary does not mean the underlying architecture no longer matters.
Especially with PVS/MCSIO + WEM + non-persistent VDAs, I would therefore include the WEM cache and its persistence in a health check again.
Incidentally, I have not yet found a public Citrix bug ID explicitly linking the current behavior to WEM 2507 CU1. For now, it remains exactly that:
Finding in the Wild. 🔎
