changed vmdk size, how to resume replication?

[doomi]

Hi

the customer resized a VMDK of the source VM. the backup just did a new full transfer of the VMDK and backed up normally after that.
but the replication of this VM now gets the error: invalid snapshot configuration detected.

how can I resume replication without doing a complete retransfer?
can I delete the destination snapshots, manually resize the destination VMDK to the same size and then seed?

[doomi]

interestingly, the customer did resize another VM, which worked just fine:
04.04.2016 22:45:48 :: VM disk size changed since last sync, deleting all restore points

is this a new feature in v9, that it handles VMDK resizes better in replications?
because the first VMDK resize happend in v8 just before the upgrade and fails now with the "invalid snapshot configuration" error.

[veremin]

is this a new feature in v9, that it handles VMDK resizes better in replications?

Nope, the said behaviour of handling replica disk resize has existed since version 8, so, I'm not sure why you originally got the said error. Thanks.

[doomi]

I had the same "invalid snapshot" error already once, when the customer shrinked a vmdk and had to do a complete retransfer.
that's why I'm interested in a way to fix it with seeding or at least keep all other vmdks.

[doomi]

ok, I'm trying to fix it as of the following KB:
https://www.veeam.com/kb1773

I will let you know tomorrow how it worked out.

[skrause]

I had the same "invalid snapshot" error already once, when the customer shrinked a vmdk and had to do a complete retransfer.
that's why I'm interested in a way to fix it with seeding or at least keep all other vmdks.

Since shrinking a VMDK requires quite a bit of manual voodoo and could possibly remove parts of the "disk" that are being referred to in the snapshot, I can see how an invalid snapshot error might occur.

Veeam will handle increasing the size of VMDKs fine in most cases in my experience.

[doomi]

it worked as described in this KB:
https://www.veeam.com/kb1773

I had troubles getting rid of alle the old VMware snapshots. had to try a few times with adding and removing the VMDKs, creating/deleteing/consolidate snapshots.
but in the end I could resume the replication (which took a VERY long time for fingerprinting, because it's a 700GB VMDK).