Deduplication ratio on Avamar

Compare Totalbytes colum to modsend culum, it will give you dedupe ratio.

Regards

Lukas

Hi Lucky85,

is this information available for the whole avamar system?

In the reports I only can see these entries for each single client.

Of course I can export this and calculate the ration by myself, but it would be easier to see it in the reports in one step.

No as far as I know, you have to calculate it, use dpn summary report from the beginning and import it to excel…

Regards

Luksa

Hi Lucky85,

Thanks for your reply.

I calculated the dedup ratio using the columns you said and I am seeing a number between 500 to 1000 for filecluster servers. Now I understand that the number will be high but numbers upwards of 500???

Does this look right? or Am I doing something wrong?

Seems to be wrong, typical dedupe ratio is 20-50x.Check if you also counts initial backups…

Regards

Lukas

rmartin@ipm.es wrote: Hello, I have done the same calculation (by summing the TotalBytes column, summing the ModSent column and then dividing the TotalBytes sum by the ModSent sum.), my resullts are 108:1, is it possible? Thanks

108:1 seems too high to me.

Are you certain that the DPN Summary covers the entire history of the grid? If the initial backup for each client is not included in the report, the calculated dedupe ratio will be too high.

Hello,

I have done the same calculation (by summing the TotalBytes column, summing the ModSent column and then dividing the TotalBytes sum by the ModSent sum.), my resullts are 108:1, is it possible?

Thanks

Hi,

Are you certain that the DPN Summary covers the entire history of the grid?. yes

I will perform calculations again. Is there another way to calculate this?

Thanks

If the system has been in production for a long time, the activity records for the initial backups may have been pruned from the database, in which case they won't appear in the DPN Summary report. In the default configuration, entries older than 1 year will be pruned to keep the database at a reasonable size. This would be good to check.

If the intials for any client were taken more than a year ago, you will have to do some fiddling with the report or the calculated de-dupe ratio will be much higher than the real ratio. For any client older than a year, duplicating the oldest entry and copying the TotalBytes value into the ModSent column for that row will serve as an estimate of the initial size of the backup. This won't be perfect because it doesn't consider data that has been de-duplicated because it appears on multiple clients but it should be close enough for most purposes.

Hi ianderson,

I have calculated the dpnsummary from day one until today and we are getting Value of 82,7%. Could this be real.

Since they one all Backup in Totalbytes are ~853TB and we have a Single Node with 7,8TB.

Can this report be real?

Thanks and regards,

One thing I did forget to mention is that the DPN Summary report will include backups that have expired from the system. These will need to be excluded from the calculations.

Okay, but how can i find out which backup have already been expired, as there is no such column within this table?

I guess creating a customized view for tables could solve this issue but it need some scripting effort for this.

maybe somebody already did something similar?

You can retrieve this information from the v_activities view in the Administrator Server's database (including the expiry dates and times). The System Administration guide explains how you can set up an ODBC connection to the database. From there you can use third party tools (spreadsheets, database packages, reporting packages, etc.) to query the views directly.

I'm not aware of any script available that does this "out of the box" but if you develop a script, please feel free to share it on the support forum.