Reproduced below is an output of the command that shows that the value of "Capacity used" is greater than "Bytes protected"
root@avamar://#: mccli server show-prop
0,23000,CLI command completed successfully.
State Full Access
Active sessions 0
Total capacity 23.3 TB
Capacity used 20.8 TB
Server utilization 89.2%
Bytes protected 16.1 TB
But, in fact, the the value of "Capacity used" should be much less. The license allows to have 11TB of protected data max.
The current status of the OS capacity is 65%. Is there any way to reduce the used capacity and server utilization as a result?
Solved! Go to Solution.
Well, apparently I figured out how it works.
I see that status.dpn displays a %Full value of ~ 57%. This number can never exceed the value equal to the disk read-only parameter which should never be higher than 65%).
Using the values above I have 57/65 ~ 89%
So, it seems that the situation I have is normal. Could anybody confirm this assumption please?
If so, %Full value of 57% is represented the proportion of actual physical space, i.e., used space is 23,3 TB * 57% = 13,3 TB
How to figure out what part of this used space is used for client backups and what for the rest?
The 'Bytes Protected' is not very accurate. It is not used to caculate how much data was backed up precisely. The detail is explained in this KB.
We've asked engineering to remove this from GUI as it is so confusing. But it could give customer an estimated number, so that it is stil there.
Here's an alternative URL for the KB article that Beta was referencing. It will persist over time even if a newer draft of the article is published
Hi Zhou, Nicholas,
Thank you for the links provided. I understand that 'Bytes protected' need not correspond with the Total capacity and Capacity used values. That's right, they are calculating different things.
Now I'm trying to understand what's the possible reason Capacity used has achieved extremely high value of 20,8TB, given that the license allows to have not more than 11TB of protected data?
Well, it is not easy to find out what data is on the Avamar server as all data is deduplicated. From my experience, we can start from DPN SUMMARY report. Exporting to an excel file and using excel features to find out what clients contribute much new data.
There are also other possibilities, for example, the GC doesn't clean the old data. We will need to do a heath check on this grid.
The license doesn't audit the amount of data on a system.. it's more like a binary switch which allows it to function when present and valid. You can discuss the licensing amount with a Dell account manager.
Regarding why the capacity grew so much, as Beta suggested you could make use of the DPN Summary report (refer to the Avamar Reports Guide) or perhaps a more friendly tool could be the capacity script
Hi Zhou, Nicholas,
Health Check revealed the following:
Errors to be reviewed:
ERROR: Node 0.s selected *127.127.1.0 as the primary time server!
It's no big deal.
I have no more questions regarding the server utilization value. I will use capacity.sh, dpn summary report in order to revise the datasets. But, I still don't understand: if 13,3TB is used for the backups, used capacity value is 20,8TB, what the remaining 7,5TB is used for? OS, CP's, logs? 7,5TB that's a lot. I read that 20% (~4,7TB in my case) is reserved for checkpoint overhead. Is it true?
Bytes protected is not listing capacity used by backups. Bytes protected = The sum of the amount of space required to restore the largest single backup on each client which currently has a backup on the Avamar system. Used capacity is the amount of space used by all backup data after compression and RAIN. I don't believe it includes checkpoints. See previously linked KB for more details.
As mentioned in the KB 4.7 TB in your case could be:
- RAIN redundancy
- Data due to Retention
Not specifically mentioned in the KB but my understanding is the following would also not be counted in bytes protected:
- retired clients
- Any data replicated to grid (Clients in the REPLICATE domain)
The OS, CPs usage doesn't count in the used capacity. There might be other backup data, or expired data is not cleaned.
If you still have problem, feel free to open a support case. That will be more effective to get it figured.