Anyone using Citrix "RadeCache" - disk space alerts

[Jarv]

Just wondered if anyone uses or has experience of Citrix RadeCache on VM's that could possibly explain to me to what is going with "Virtual machine disk space anayalysis" alerts from Veeam. I dont think this is a Veeam issue but perhaps just how the above works. Somebody has tried to explain to me that its a folder where Citrix stores its application cache and it saves it as a virtual disk file.

The disk partition that is being breached is listed as c:\program filex (x86)\Citrix\RadeCache\"somerandomnumber"\-47.87MB

Thanks

[Vitaliy S.]

Hi Steve,

Can you please clarify a bit more on what you observe? Do you see alerts from this monitor with the disk partition you've specified above? Does it happen on particular time or it is just random?

Thanks!

[Jarv]

Hi, and thanks for the reply. Like I say I dont see this as a "Veaam problem" but could do with understanding how I may work round it. Unfortunatly Citrix left me behind long ago but this is the information I got back from the SME -

"radecache is part of citrix streamed applications, and is a folder where it stores the application cache.

Basically, I prepackage the application and save it on our file server. I then publish the application to this location. I use a feature of streaming to save the application package as a virtual disk file.
When the application is launched,Citrix creates a junction point to this .vhd file. This fools the operating system into thinking the application is stored locally, but it is actually saved on the file server."

In Veeam I just took a look at the Performance views, disk, freeMB on vdisks for the server involved. It shows up as 7 instances. One of these is drive C (which is the only disk in the server). It then shows 6 instances as follows -
c:\program files (x86)\citrix\radecache\"some sort of random GUID"\

I also just took a look at the server involved. In this directory I can only see 2 of the above. This would seem to suggest the "random GUID" directory creates and deletes and changes. I guess this would cause a problem even if I could override for a disk object as it would change. As it happens these servers are new and have a lot of disk space. I cant see them running out so may have to just create a group of Citrix servers and override?

[sergey.g]

Hello Steve,

So, as far as I understand the problem exists when our management packs discovers this attached .VHDs. For the OS this looks like a regular local disk, looks like these VHD files have only space necessary to store the application and less than 100MB (which is a threshold in our MP) and our monitor immediately alerts about the disk with lack of free space.

However there is a workaround. We have an overridable parameter for the "Virtual Machine Disk Space Analysis" - Exclude Disks. Your situation is complicated because you have disk which are attached to the folder and with constantly changing name. But our MP allows regexp wildcards in this parameter, so if you add "|c:\\program files (x86)\\citrix\\radecache\\.*" to the default overrides (|("pipe") is a separator for multiple disk labels, ".*" in regexp means any symbol repeated infinite times) - our monitor will exclude these RedeCache drives from analysis.

There is also another workaround - you can override the threshold, if these VHDs have let's say 80 MB of free space - you can set a threshold to 70 MB.

Let me know if you have any additional questions.

Thanks.

[Jarv]

Hi Sergey, just got round to testing this in my SCOM and also live environment and cant get the exclude disks to work..

I chose the "Veeam VMware: Virtual Machine Disk Space Analysis" monitor and did an override for all objects of class - VmWare Virtual Machine

The Exclude Disks had the following in it -
A:\\|B:\\|/boot

I changed this to -
A:\\|B:\\|/boot|c:\\program files (x86)\\citrix\\radecache\\.*

However the alerts have continued. I just went one one of the servers and cut and pasted the directory exactly from that. Aside from capitals it looks exactly the same -
C:\Program Files (x86)\Citrix\RadeCache

Is it perhaps the .* at the end. It needs to be *.* or something perhaps?? thanks

[sergey.g]

Hi Steve,

I'm sorry, but I'm afraid that's my fault - I didn't do a test with "(x86)" part.
I had to re-confirm the behavior for this override. This parameter uses Regular Expressions, so if you want to encode "c:\program files (x86)\citrix\radecache\", I beleive it should be "c:\\program files \(x86\)\\citrix\\radecache\\.*" becuase in regexp "()" brackets are special symbols and should be encoded with backslash '\' as well as backslash itself - that's why we have double-backslash everywhere.

I was abel to check in the lab that it should work - let me know if you still experience issues with creating this override.
Thanks.

[Jarv]

Thanks for the reply Sergey I didnt check back on this until now as I just did a temp complete override for disk space as the servers had loads of space anyway and "unlikely to fill". I have just done this override in test so should know if its worked in a couple of days.

[Jarv]

I had to re-confirm the behavior for this override. This parameter uses Regular Expressions, so if you want to encode "c:\program files (x86)\citrix\radecache\", I beleive it should be "c:\\program files \(x86\)\\citrix\\radecache\\.*" becuase in regexp "()" brackets are special symbols and should be encoded with backslash '\' as well as backslash itself - that's why we have double-backslash everywhere.

Sergey I have left this running with this override for around a week now and just to let you know it does appear to have "fixed" the issue.
thanks

[sergey.g]

Hello Steve,

Thank you very much for the update, I'm really glad to hear that everything works as expected with our Management Pack.

Let us know if you have any other issues with coexistance of our Management Pack and other solutions, in Veeam we really like solving problems which uncover the flexibility of our product.

Thanks.