Hi,
I am having an issue with a couple desktop pools in Horizon View 7.13
Desktop Pool Attributes Automated, Linked Clone, Dedicated
I have confirmed the golden image exists within vCenter and when I browse to the golden image in the desktop pool settings and click submit, it does not save the settings and the image field remains blank.
Error message: Could not find golden image in the vCenter Server.
I have tested several other images for this pool and still get the same issue. It seems to be related to specific desktop pools, I noticed 2 desktop pools with this issue. Also we have other pools with other golden images from the same vCenter and they all work as normal.
The pool was created a while ago update the golden image every so often, so now I cannot add a new golden image to provision new machines in the pool.
Has anyone came across this before?
Thanks
Same here, it won't let me choose the image again, the line stays blank
This issue can occur if the parent vm was deleted/moved from/in vCenter. It is a known issue and a fix will be released as 7.13.1.
The issue does not occur in the old flash based flex admin.
I too am having the same issue. Do you know when 7.13.1 will be released?
Thanks,
This issue has been around since the HTML client was introduced. I have an old bug post from mid-2020 about it but nothing happened
Check whether deleted/Missing VDI present in that pool if so remove it and then try to map the golden image.
I has the same issue and now i could able to map the golden image
I'm encountering this issue right now - I even upgraded composer and one of our connection servers to 7.13.1 (will upgrade the other connection server afterhours tonight) but I still can't change the gold image/snapshots on the affected pools. Did anyone ever find a resolution for this issue? This is a big deal - it worked in Horizon Administrator.....sigh.
I have an old gold image that still has Flash installed on it, and I used that to access the Horizon Flash management page
Wonderful.....hmm....I'll try and see if we have any old machines floating around running flash, but I highly doubt it.
Actually, I take that back - one of our network engineers has an older server with it - phew. I'll just use that to fix this. @VMware <-- fix this. It's ridiculous.
Are you moving the image in vcenter or anything? Personally after suffering issues with that even in 6 versions I never change anything on the vcenter side thats in use, it uses a little extra storage but I clone the parent image to seperate ones per pool, and if I need to make a change I clone that and use that for the pool and delete the older one. That way your unlikely to run into these issues, even though its was supposed to be "fixed"
For me, it happened when I just renamed the gold image in vcenter. In the Horizon Flex management page, it would simply pop up with an error message saying it couldn't locate the image, and I click OK to find it again. In Horizon HTML5, no message pops up and that image field is just greyed out.
Yeah that happens because horizon saves the state of vcenter in the associated ldap database on the connection server, the same thing if you rename anything else. Horizon, at least it used to, doesn't actively check for updates every few minutes. It expects things to be the same, I'm assuming the "fix", is to handle that inconsistency better, but if we don't make only vcenter side changes that aren't 100% necessary this won't happen regardless.
@sjesse wrote:don't make only vcenter side changes that aren't 100% necessary this won't happen regardless.
Did VMWare get bought out by Apple? That sounds a lot like the "you're holding it wrong" excuse 😉. Look, obviously it happens, for one reason or another. Maybe vCenter is also used by another team or someone renames the wrong VM? It's not something that's extremely obvious, like deleting the image for example.
It shouldn't cause an entire pool to become irrevocably broken (unless you have access to Adobe Flash). To me, that's an actual bug, not a "you're doing it wrong" issue.
FYI, I don't work for vmware, I'm just trying to share what works for us and I haven't had these issues in awhil3e
Ah ok sorry, your "mod" and "Leadership" tags look like you represent VMWare here
Also this is closer to working directly in esxi, when vcenter is managing it, vcenter doesn't always handle that gracefully. I'd like to work as your thinking, but I'm also responsible for preventing issues in our environment so I need to protect us as best I can.
No worries, with the forum software change its a bit harder too. Usually know they have (employee) in there name directly. I just respond alot here since I like to help, and I volunteer, most mods do but there are some mods that are vmware employees too.
All fixed now, thanks to an old server + Pale Moon browser + an Internet Archive version of Adobe Flash (we keep this server powered off and only power it on when absolutely needed).
Yeah late to the game - but got the issue anyway today 🙂
Another workaround that may help is to:
If you dont like the old name, because it may have an old date in it, just side step by doing the above, then:
Yes it is a bit convulted, but the only option if you dont have Flash and you dont want to remake your pool from scratch...
OK for some reason this works only on some pools - had to dig up an old flash MSI installer, and put Edge in IE mode, then i could fire up the FlexAdmin interface and fix the issue 🙂