DP for HyperV active data question

[itdcdu]

Hi

TSM server 7.1.1, DP for HyperV 7.1.6..

Our organisation is about to switch Hypervisor from VMWare to HyperV and as POC we have are using DP for HyperV to backup a single host with 5 Vm’s.

To alleviate slow restore times that we saw with VMWare I have added an Active Data Pool for the HV primary pool on a SAN attached disk.

I ran an export node preview to calculate the amount of data to be copied to the ADP, this amounted to about 65% of occupancy of the primary pool, however, when I ran copy activedata it copied the entire contents of the pool to the ADP, that was a couple of weeks ago and the ADP continues to mirror the occupancy of the primary, same amount of data and files.

Does anybody have any sort of explanation for this behaviour? I’ve searched just about every online source available and this sort of query is not covered by our support agreement with IBM, are active data pools supported with TSM for VE??

 

[marclant]

I'm not sure, it's possible all the VE data is stored as active.

Check:
select state,count(*) from backups where node_name='NODE_NAME' group by state

 

[itdcdu]

Hi

Thanks for the reply, I ran the select as suggested;

STATE Unnamed[2]

----------------- ------------

ACTIVE_VERSION 62790

INACTIVE_VERSION 330


So there are inactive versions, not very many admittedly, can you explain how VM data is expired, I know that VEREXISTS and RETEXTRA are honoured, what are the recommended MGMT class retention settings?

I also previewed another export node active data only;

11/29/2016 11:26:11 ANR0986I Process 1091 for EXPORT NODE running in the

BACKGROUND processed 18,593 items for a total of

234,970,860,437 bytes with a completion state of SUCCESS

at 11:26:11. (SESSION: 117088, PROCESS: 1091)

So around 230GB, the node has 332GB total occupancy, so that would suggest 30% is inactive?

 

[marclant]

Q9, Q10 and Q11 should answer that: https://www.ibm.com/developerworks/...on for Microsoft Hyper-V?section=Q9 retention