AWS Issues and Workarounds

Question

What is the best practice for shutting down clustered VMs in AWS?

Answer

If shutting down the primary source node, all cluster roles depending on a SIOS DataKeeper Volume Resource should be placed in the Offline state. Also make sure all mirrors are in the mirroring state prior to shutting down any VM. The node shutdown order does not matter as long as the previous steps have been taken.

Ephemeral Storage message displayed in DK GUI when the Bitmap location is NOT ephemeral storage

Description

ISSUE: After the upgrade to 8.8.0 (or later), a red message in the DK GUI stated that the Ephemeral storage configuration was not correct.

Version: DataKeeper Cluster Edition 8.8.0

Example Configurations:

• DK running in AWS and has the BitmapBaseDir setting was left with the default value (on C:\ drive) instead of pointing to a location of ephemeral storage.
• No ephemeral storage is used for the bitmap.
• DK running in AWS, but ephemeral storage is not configured properly through EC2Launch
  
  

Workaround

If bitmaps are not set to ephemeral disks, then you will receive an error message within the GUI.

For information on DataKeeper Intent Log files on AWS Ephemeral Storage, see HERE.

Additional Information

For information on changing the bitmap, refer to Relocation of the Intent Log.

Please see Upgrading SIOS DataKeeper Cluster Edition for latest upgrading instructions.