L2 Cache gone

[craggy]

We have a big issue where we updated to latest version ov V8 (as of 18/4/15) and both our 28TB volumes have lost their L2 caches resulting in data corruption on out VMware datastores.

Why did this happen?

How can we re-add the L2 caches without having to remove the devices from the targets causing the need to re-signature luns in VMware?

[Oles (staff)]

Hello Craggy,

To add L2 cache, please follow next steps:

1. Create 1 device for L2 cache of necessary size. Do not attach it to any targets. Location of L2 device’s .img and .swdsk files should be path to your SSD drives’ array

2. Stop StarWind service

3. Open the corresponding .swdsk file of L2 device and copy the section that says this:

<storage id="3" name="My Computer\D\img\L2-Storage11.swdsk" type="device" lun="0">
<interval size="XXXXX" units="MB"/>
</storage>
Note: paths, size and storage id will differ in your case.

4. Open the corresponding .swdsk file of your main device and paste this section there between <storages>and </storages> tags.

5. Now add the following section keeping “storage_ref id” the same as “storage id” above (in this case it is 3):

<cache type="write-through" size="XXXXX" units="MB" level="2">
<storage_ref id="3"/>
</cache>
This should be added in-between <caching> and </caching> tags.

6. Start StarWind service

7. Now L2 device should actually disappear from SW console.

To understand why did it disappear, please submit a support ticket and please provide me with next information:

May I ask you to provide me with next information:
· StarWind logs from all SAN boxes
· Windows Application and System logs (in *.csv format) from all SAN boxes
· Detailed network diagram of SAN system
· Description of the actions that were performed before/at the time of the issue
· Approximate time frames when the issue happened
 
I`d appreciate if you`ll separate the logs from different servers into the different folders
 
 
Thank you

[craggy]

Thanks for the response.

An observation is that surely this should be a function that is available from the GUI as the ability to add/remove/migrate L2 cache to another pool as SSDs fail etc. shouldn't require downtime and an outage of the storage served from SW.

The thing is that my previous SW8 version had the L2 caches and once we upgraded they disappeared even though the data for them remains in the config files and the cache files remain on the SSD.
When I check the Resource Monitor in Windows the SSD holding the L2 cache files is not being used whatsoever.

[Oles (staff)]

Craggy,

I differently agree with implementation of this process into GUI, we are currently working on it.
About the cache that has dissipated, have you tried to add it back using the guidance?

[pstoianov]

craggy, According to the documentation, even if SSD completely failed, no issue is expected with the storage as is L2 in WT (read-only) mode.

What was your L2 configuration mode? WT or WB?
I am planing to use in production L2 cache in WT mode, but still checking if anyone facing issues with L2 in WT mode.

[Oles (staff)]

Hi guys!

If you have any questions left for me, please let me know, thank you.