CBT Reset\Re-seed replica

[kjc3303]

I had to reset CBT on one of my VMs as per KB due the cbt soap error, not an issue with my backup jobs as full disks reread and cbt recreated in a matter of minutes and all working fine now . My replica job is crawling along doing a full read of the disk and is going to take days.

2 Questions

1 Is it normal for it to take so long to do a full read over a slow link (10mb)? Bottleneck is network.
2 Can I just take a copy of my backup over to DR site and reseed? without having to remove the replica VM and restore points?

Thanks

Kev

[veremin]

Can I ask you where replica metadata is being stored?

Thanks.

[kjc3303]

at the source(Production) site

Thanks

[veremin]

I’m wondering whether the whole VM image is only reread or it’s also transferred as a whole via WAN connection to the remote site.

Furthermore, how many proxy servers are utilized for this replication job and where they are located?

Thanks.

[kjc3303]

Vladimir, thanks

I have just been looking into the stats a bit more and I think you are correct. So far the job has transferred 19.2GB of data (There has not been this amount of change since last Replication) @ 1.8x compression and so far the read data is 34.6. This ties in with what you have suggested above.

I have cancelled the job and will either remap existing replica or reseed from backup, what would you suggest is the best tactic for this

Kev

Thanks

[kjc3303]

Proxy servers are correct - source proxy at Production site and Target proxy at DR site

[veremin]

Please be aware in case of you using replica mapping functionality all previous restore points will be lost.

During the first run, only differences between VMs will be calculated and transferred. These differences later will be used to synchronize the state of VM in DR site and the original one in Production site.

After successful synchronization, there will be only two restore points. One of which is the latest state of the mapped VM (target VM), another is the latest state of the original VM on the source host.

Hope this helps.
Thanks.

[kjc3303]

Thanks for the detailed explanation Vladimir, this will be OK as we only keep 7 restore points and replicate every 2 hours so will build up quickly and is certainly better than waiting for two\three days with no replication.

Thanks again

Kev