2011-08-09 06:40 AM
We are a relatively new NetApp customer and working through our backup policies to find some efficiences with the addition of NetApp snapshots. We are a fibre channel shop and started allocating 800GB LUN's to our VMware ESX hosts inside 1TB volumes. Each 800GB LUN holds between 4 and 8 virtual machines. Some of our daily NetApp snapshots are around 3GB while others are closer to 50GB. While we have started seperating our our SQL virtual machines to LUN's that are only snapshotted once a month (we are using a seperate process to capture the nightly SQL dump), but some of our nonSQL boxes most have a high rate of change. How do you identify which VMs/files are changing when there are multiple VM's on a LUN?
Thanks in advance...
2011-08-10 07:05 AM
Thanks for your reply Radek. It is a related thread, but it doesn't really answer my main question: How do you identify which VMs/files are changing when there are multiple VM's on a LUN?
If I have a LUN with 8 VM's with 2 VMDKs each....how can I tell which one is causing the bloated snaphots?
2011-08-18 01:03 PM
This is a tough thing to do from the storage end. The only way I can think to do it would be to open VMware snapshots on the VMs in one of the LUNs and monitor (Closely!) the VMware snapshot growth. You obviously don't want to leave this for long but you should be able to tell in 30 min or so which VM is generating the change. If it isn't obvious, perhaps one of the VMs is doing a backup or a dump to disk off hours which is driving up the snapshot size. In that case you want to schedule hourly NetApp snapshots to determine when the grwoth occurs then inspect the VMs to see if they are doing anything odd at that time.