wooody_collins
6 Indium

New to RP4VM...

Hello!

I am fairly new to recoverpoint for vm and so far my experience with the product has been less than stellar.  Im not entirely sure that isn't due to lack of knowledge along with complexity of our environment.  Could one of you more experienced users point me in the direction for some type of better documentation or web postings that give a better direction on troubleshooting and working with the product?  I have read through all the PDF's that have been sent to me by EMC and started using various CLI commands but am still very confused with the product and especially frustrated with the user interface (just about completely useless if you ask me).

The company I work for has purchased product licensing with support but im in a limbo state with my support account waiting for them to approve my account with the site ID information so I can open support cases and was hoping to get a better understanding on the product in the mean time.

If anyone is curious as to some of the issue's im running into, we basicly have 2 sites with 2 vCenter servers running in linked mode.  I've stood up 2 clusters (one in each site) and have joined them.  I have presented a journal volume at each site that is seen by all the nodes in the ESX cluster on each site and for the first couple of protections I did on a couple of VMs they init'd correctly and eventually went into an active state.  The data transfers happened as I could see it flowing through my performance charts etc so I was very happy.

Then things have gone down hill.  I unprotected those VMs to make sure they worked (they did).  But now any time I try and protect a VM the consistency group goes into a "Paused by system" state and eventually makes it to error.  I've managed to track down how to use the CLI commands to get log information but really have almost no understanding of things involved with the tech stack when it says things like "group wont run on any box" , or "can't maintain history - not paused on snapshot and  user volume problem - Volume issue" or "The copy marked all of its replication volumes as dirty".

From what I can tell the infrastructure setup isn't outside any scope of problems.  The management network should probably be isolated, but the documentation doesnt state any oddities on requirements when it comes to LUN presentation for the source and destination datastores needed for the copies etc (as we do have a very odd setup currently that is being worked through to resolve).

Any help would be appreciated.

Thanks

Labels (1)
Tags (2)
0 Kudos
14 Replies
8 Krypton

Re: New to RP4VM...

There are a couple of known issues that may explain this behaviour and these can be found under RecoverPoint for Virtual Machines -> Knowledgebase on EMC Online Support.

What version are you running?

Regards,

Rich Forshaw

Consultant Corporate Systems Engineer - RecoverPoint & VPLEX (EMEA)

Data Protection and Availability Solutions

EMC Europe Limited

Mobile: 44 (0) 7730 781169<tel:44%20(0)%207730%20781169>

E-mail: richard.forshaw@emc.com<mailto:richard.forshaw@emc.com>

Twitter: @rw4shaw

0 Kudos
wooody_collins
6 Indium

Re: New to RP4VM...

4.3.1(h.192)

0 Kudos
wooody_collins
6 Indium

Re: New to RP4VM...

Also, any chance you know a solution to this error:
"The copy marked all of its replication volumes as dirty."

Thats the event that keeps happening in the event logs of the RPA's and so far my searching in the knowledgebase has turned up 0 luck.

0 Kudos
wooody_collins
6 Indium

Re: New to RP4VM...

Alright, so I have narrowed the trouble down.  It seems that for whatever reason, if the VM is living on a LUN that spans multiple storage groups (I believe that is the correct term, but i could be wrong) that the process never makes it to the init procedure.  It creates the journal content and deploys out the VMs correctly and even boots the shadow machine.  Then goes to a "paused by system" state and eventually an "error" state and no transfer occurs.  If the VM resides on a LUN that doesnt span multiple storage groups everything works as expected.

Another thing to mention is that in the eventlogs there isn't anything reported.  It gets to a step that says something about the copies are marked as dirty and on the next sync will be a full sweep but never takes the next step.

I have no idea why the storage presentation would matter.

0 Kudos
8 Krypton

Re: New to RP4VM...

I would recommend an non-disruptive upgrade to 4.3 SP1 P2.

Regards,

Rich Forshaw

Consultant Corporate Systems Engineer - RecoverPoint & VPLEX (EMEA)

Data Protection and Availability Solutions

EMC Europe Limited

Mobile: 44 (0) 7730 781169<tel:44%20(0)%207730%20781169>

E-mail: richard.forshaw@emc.com<mailto:richard.forshaw@emc.com>

Twitter: @rw4shaw

0 Kudos
8 Krypton

Re: New to RP4VM...

I will check to see if this bug related but what is telling you in technical terms is what would is required during a full sweep or alternatively what happens if you ran the set_markers CLI command.

All volumes are dirty and require syncing to reach a clean status.

Regards,

Rich Forshaw

Consultant Corporate Systems Engineer - RecoverPoint & VPLEX (EMEA)

Data Protection and Availability Solutions

EMC Europe Limited

Mobile: 44 (0) 7730 781169<tel:44%20(0)%207730%20781169>

E-mail: richard.forshaw@emc.com<mailto:richard.forshaw@emc.com>

Twitter: @rw4shaw

0 Kudos
wooody_collins
6 Indium

Re: New to RP4VM...

As far as I can tell in the event logs , every machine that I have protected states that dirty comment prior to the init of the protection so I believe the procedures that occur upon first protecting a VM are occurring correctly so I doubt it's a bug.  I have a feeling there is something else in the environment causing the problem.  With the way that things are configured here, the volumes I am having issue's getting VM to replicate from, 2 of the 3 hosts that are presented those volumes are having splitter errors.  I have a sneaky suspicion that either the networking team here has mis-configured the ports used for the iscsi connections or our SAN team has used IP's on the network segment and not mentioned to anyone and there are IP conflicts.

Sadly I wont be able to troubleshoot those problems till next week when I can get everyone together.

So far thank you for your help, I will look into upgrading to the newer version you mention.  I just stood the environment up on Friday (like 3 days ago) from software given to me by our SAN team and if there already is a new version out I would prefer to use it.

0 Kudos
8 Krypton

Re: New to RP4VM...

The error would indicate that there is a splitter related error on the basis of the dirty volume reference. As part of a full init process all volumes are dirty. If this continues to happen there is an underlying issue.

In would check the iSCSI config each of the ESX hosts.

Regards,

Rich Forshaw

Consultant Corporate Systems Engineer - RecoverPoint & VPLEX (EMEA)

Data Protection and Availability Solutions

EMC Europe Limited

Mobile: 44 (0) 7730 781169<tel:44%20(0)%207730%20781169>

E-mail: richard.forshaw@emc.com<mailto:richard.forshaw@emc.com>

Twitter: @rw4shaw

0 Kudos
wooody_collins
6 Indium

Re: New to RP4VM...

I wanted to say thank you for all your help.  Turns out my problems were all related to networking.  Once I resolved some port configuration issues as well as address conflicts things seem to be working much better.  The next thing I am going to try and do is hunt down documentation on how to create policy templates.  Any idea were that might be found at? I've been searching through the administration guide and it talks about loading templates but never really discusses how to create a new one.

0 Kudos