Workload Balancer (WLB) Appliance Disk Space

[sortola27]

Hi folks,

 

I have a Citrix Workload Balancer (WLB) appliance that is currently out of disk space.  From what I've been able to gather it's pretty much using all default settings, with a 30GB disk right now that's 100% full.

 

We currently have about 250 VMs. According to the  WLB space formula in the admin doc:
 

Total estimated disk size = ( ( number of days * average disk usage ) + base disk usage ) * grooming multiplier

Using, say, 30 days and .375 as the disk usage, we come out to only needing 17GB of disc space.  So  not sure why the disk would be full, unless it's storing more than 30 days of info by default.

Anybody know what the default "number of days" value WLB might be using and where we can change that?   We just want to see if there anything we can adjust to free up space before we extend the disc size any further.

 

Thanks,

[Boby John1709155536]

Try this ?

https://support.citrix.com/article/CTX330977 

 

This update fixes the following issues:

The Workload Balancing database can grow very fast and fill the disk.

 

[Jose Pablo Benavides]

You might want to check logs as well.

I have seen disk being filled by WLB logs.

[sortola27]

Did a little folder size digging and found this folder is 24G is size:

/var/lib/pgsql/data/base/16384
 

[Jose Pablo Benavides]

What version of WLB?

[Jose Pablo Benavides]

Also what version of Hypervisor are you using?

[Jose Pablo Benavides]

There is a known issue in WLB 8.2.0.

The database grooming service is not working as expected.

 

You can double check and customize GroomingRequiredMinimumDiskSizeInMB. From the document: https://docs.citrix.com/en-us/citrix-hypervisor/wlb/administer.html  (Control database grooming Section).

GroomingRequiredMinimumDiskSizeInMB controls the minimum amount of free space left in the virtual disk assigned to the Workload Balancing virtual appliance. When the data in the virtual disk grows to the point where there is only minimum disk size left on the virtual disk, Workload Balancing begins to trigger database grooming.

 

The issue on WLB 8.2.0 is that the grooming is triggered but its failing.

 

There is already a fix for this issue to be released in WLB 8.2.2. I dont have a ETA for this release.

 

 

[sortola27]

Looks like WLB is v8.2.1

 

And I if I'm looking at the correct info (new to a lot of this), Hypervisor is 8.2

[Jose Pablo Benavides]

I'm afraid WLB 8.2.1 still has the issue.

[sortola27]

Ah,ok.  Any suggested workarounds for now?  Seems like we could extend the disk but sounds like it would eventually fill up again if Grooming is cleaning stuff out.

 

Thanks for all the info.

[Jose Pablo Benavides]

2 minutes ago, sortola27 said:

Ah,ok.  Any suggested workarounds for now?  Seems like we could extend the disk but sounds like it would eventually fill up again if Grooming is cleaning stuff out.

 

Please get with Support.

 

Yeah, extending the disk its just a temporary  solution, not recommended as it will fill up again.

[Tobias Kreidl]

Looks like you can resize the disk that holds the WLB data; see partway down in this document: https://docs.citrix.com/en-us/citrix-hypervisor/wlb/administer.html

[sortola27]

We have 8.2.2 running right now and so far so good.  Thanks for the link.