I had an excellent course in Melbourne last week, not many students but all good people.
One student was a VMware staff member in the consulting team here in Australia, despite being in more of a management role there was some very good material from his experiences.
1. Six core Dunnington
One was this note about a configuration change to make if you have Intel’s new “Dunnington” 6 core CPUs in your new ESX servers. Shouldn’t they be called hex-core? maybe a hex is not something you want on your server.
2. VCB silly defaults
When VCB 1.1 was released one of the nice new features announced was that a VCB backup would delete any leftover VCB snapshots it finds when it starts a backup, rather than failing the new backup as previous versions did. Unfortunately the setting that controls this is disabled by default.
look in C:\Program Files\VMware\VMware Consolidated Backup Framework\config\config.js on your VCB proxy server and if they aren’t set to sensible values reconfigure the settings for PREEXISTING_MOUNTPOINT, PREEXISTING_VCB_SNAPSHOT, MAX_RETRIES and BACKOFF_TIME. The defaults are fail, fail, 0 and 10 which means that existing VCB snapshots or mounts on a VM will cause new VCB backups of the VM to fail. Changing to delete, delete, 3 and 60 will cause the existing mount points and snapshots to be deleted (changes written into parent disk) and will allow up to 4 minutes for these to complete.
There were also lots of obscure issues with Converter and old NT 4.0 machines, which can be summarised as you’ll have “fun”.
© 2009, Alastair. All rights reserved.