Replicator product utilizes snapsure checkpoints (2) to detect new and changed blocks and send them over to the target file system.  What is confusing to users ?

Yes. All of our production file systems are replicated to a disaster recovery site.

Could one of you be so kind as to provide a quick explanation of replications impact on checkpoints?

Also, this is quite confusing to end users trying to restore files for various reasons. Any suggestions on how to deal with that? Can I mask those temporary checkpoint somehow?

Thanks again!

Is the file system replicated?

You could see more if that file system is replicated to another Celerra/VNX

When you select "Restore Previous Versions" in a windows environment, it shows the available checkpoints to choose from.

It is entirely possible that they will select one of the latest checkpoints available. It is not uncommon (but hit or miss) that their selection will throw an error because that checkpoint they chose to use is no longer available. If they choose one of the older (scheduled checkpoints) to restore from, all is well with the world.

I assume it is what it is and that I will have to provide proper documentation for our service center to respond to such questions from the user community.

Thanks for your help!

The scheduled checkpoints are available.

[nasadmin@cdcnas2-cs0 ~]$ fs_ckpt homedir -l

id    ckpt_name                creation_time           inuse fullmark   total_savvol_used  ckpt_usage_on_savvol

7212  ckpt_homedir_1_001       07/09/2013-01:20:03-EDT   y   90%        5%                 1%

7217  ckpt_homedir_1_002       07/10/2013-01:20:02-EDT   y   90%        5%                 1%

7220  ckpt_homedir_1_003       07/11/2013-01:20:02-EDT   y   90%        5%                 1%

Let me give you a couple images of what a windows user would see:

As you can see, the top entries in the list are changing. In these examples, the top 2 (which I now assume are replication related checkpoints) are "transient" and the ones following are from our scheduled checkpoints.

Depending on how long it takes the user to select one of them, they either get to work with the contents of the folder or they get an error (version not available).

Thanks for following up.

Dave

Dave,

that should not be happening, if you have scheduled checkpints, they will get automatically refreshed so what users see in "previous versions" tab should be available. Can you double check that all snapshots are available ?

fs_ckpt -l

the 1:20am are you scheduled checkpoints and they always work ? It's the two replication ones that sometimes work and sometimes do not ?

Correct.

My intent here was to understand 1) what exactly am I seeing here and 2) what, if anything I can do about it?

We have recently moved a considerable number of users to our NAS environment. And, as you might imagine, they would like the ability to recover files without assistance. You might also imagine that a user looking at the screenshots I provided earlier would think "Hey, I can recover the file to a time right before I messed it up!". However, as your know, this may not be the case.

As I mentioned earlier, if there isn't much I can do about functionality at this point, I will simply have to document/train accordingly. This implementation will just appear a bit "clunky" to the users I am sure...

Dave

I have not tested extensively, but in my experience:

1) The scheduled checkpoints (taken at 1:20am) have consistently worked as I would expect.

2) The replication checkpoints sometimes work and sometimes not. I believe these work if I can catch them before they change...

Thanks Dynomox.