Chances are that your robocopy isn't working correctly.

Try using "emcopy" instead of "robocopy."

Home > Support > Product and Diagnostic Tools > Celerra Tools > Windows Migration Tools > EMCOPY

-Marty

hi rainer

i was already using the /fft option and when i check the robocopy log it has only copied a fraction of the total data over (total 170gb copied 5.3gb). I have deleted the replication and setup again and stopped the robocopy job to see what happens after the FULL COPY

if nothing is copied across when (because nothing is being updated) i will try emc copy

as there is already data there would the below emc copy config be correct

emcopy_V4.11.exe \\***ton\groups \\***cifs\groups *.* /o /s /c /r:1 /w:1 /f /sdd /de /purge /log+:C:\robocopy\EMCcopy\DatacopyGroups.txt

we are doing this on 3 other celerra's with no issue so its a bit strange

Thansk for your help

You should be able to identify from the data mover logs if Replicator is really doing a full sync (which I doubt).

If not please open a service request and let EMC support investigate

I think its more likely that your robocopy jobs are by mistake copying everything again – then Replicator being block based also has to copy that even with an incremental copy.

The savvol increasing pretty much points to that as well

Use emcopy instead

It might help to set cifs.nanoroundoff or tell robocopy to only use seconds as time granularity

Rainer

robocopy could be changing time stamps or something.

The logs or a nas_replicate –info will clearly tell you if the current transfer is a full copy or not

IF its really doing multiple full copies in a row for the same file system then PLEASE open a service request.

The support forum does not get EMC support formally involved.

Another reason for getting data more transferred than changed from the client side could be deduplicaton.

Rainer

Hi Guys

    have logged a call will let you know what happens

Paul