Arcserve® UDP Data Store Capacity Planning Tool

[b.vanhaastrecht]

Hello,

Sorry to throw in an competitor, but they actually have a tool to do backup storage calculations. Can Veeam make a tool like this: http://documentation.arcserve.com/Arcse ... _guide.htm

Planning/calculating setups with datasets less than 10TB can be estimated with educated guesses and by using the "Restore point simulator", but with larger data sets a few percentages can make a huge difference in total space required. If Veeam would have a tool like this, we can do estimated capacity planning’s.

Kind Regards,
Bastiaan

[dellock6]

Actually, we have Veeam ONE that has exactly the capabilities you need. Even with the 30 days trial, you can monitor a virtualized environment, grab info like datastore consumption (for full backups) and daily changes (for incrementals). Then you can use rps to simulate the backups.

This is what I did last time at a large bank with 1000+ VMs, compared to their guestimation we tuned numbers by almost 70 tb per day.

[b.vanhaastrecht]

True, Veeam ONE has some metrics in capacity planning, but does not do a data set analysis on dedupe ratio. With the Restore Point Calculator you need to estimate the compression/dedupe ratio while the Arcserve tool does an analysis of all to be back upped VM's and applies its dedupe hash on it and gives an exact number on required storage needed. Customers differ on dataset, exact analysis and calculation is better than estimating.

And, Veeam ONE requires an full installation. This tool is less than 10mb.

[dellock6]

Curious, how can it hashes the entire data set? It reads all production blocks? It's almost like running a full backup in terms of io consumption

[b.vanhaastrecht]

Yes, it deploys an probe in the VM (node) and analyzes the data. On a 100GB VM takes a few minutes.

You can run the tool for free, try it if you want

[dellock6]

I'm away for the week... So, it runs inside a vm? Means you need to run it inside any vm, or I'm getting it wrong?

[b.vanhaastrecht]

Yes, partly. The tool (collector) deploys the probe's to node's (Windows based VM's), each prope analyzes the data in the VM and sends the result to the collector. After each VM/probe has reported their result to the collector you can create a report with some parameters like retention and the result is backup size needed.