Enterprise Manager does not save replication settings

I would recommend opening a ticket with the support team for this.

Rob:

Did you ever get a resolution to your problem as we are experiencing the same issues.

Thanks

Ken

Which parameters are not being saved? Is there one specific parameter or is it all of them?

I may know what's happening but I need more detail to be sure.

It is all of them.  EMC has been troubleshooting but has not found any resoultion as of yet.

If you have a service request open, could you please send me a private message with the SR number? You can send me a private message from my profile.

Interesting.  I just did a new install and we are having the same issue.  I just opened an SR friday about this.

Jerry - hopefully I can save you some time if your configuration is the same as ours.  It took 5 days to finally find out that our issue was by-design.

We have two (2) avamar systems on totally different networks.  We had only allowed Port 29000 to our storage nodes of the replication target from the utility node of the source.  This implies ssl based replication.  This requires you to initially configure replication from the command line, NOT the Enterprise Manager.  Once it is initially configured from the command line, there are certain replication settings that you can changes from the Enterprise Manager, but others such as scheduled start time, need to be done via command line.

Hope this helps

Ken

Ken,

That is interesting.  When we did the install for this customer, we did not regulate replication other than configuring bond2 for replication from grid to grid.  The grids are on seperate networks, one at production data center and the other at DR data center.  The parameters that we are trying to get to save have to do with choosing a selective replication.  You can choose selective, bring up the list of clients, choose your clients then click save.  It just defaults back to Full and does not save any of the clients chosen.  Schedules work as configured but they are getting Full replication to the /Replicate domain on the destination.  I did not initially configure the Replication bond during installation as we were having some other issues happening with the install so I left it out when using dpnnetutil, but added it in after the fact.  I wonder if that has something to do with it.

Thanks for the input.  When EMC gives me a call I will see if by chance this issue is the same as yours.

Thanks Ian, I've seen this a couple times. It seems to happen more often with targets of different versions/hardware and also when the seperate replication bond is created.

Did you get a fix on this Rob?  I still have not heard back from support?  I know that both versions are the same, the only difference was that I did create the replication bond afterwards on the source server.

Hey Jerry, What versions are you running ? I know the current workaround is to manually update the cron files, I have sent a couple SR's to Ian and hopefully he can post something more meaningful.

From the discussions I've been having with the people who have posted in this thread, there seem to be a few separate issues that are getting lumped together as "Enterprise Manager does not save replication settings".

One issue is that the "dstaddr" field in the Enterprise Manager is not being updated correctly but the EM and replication are otherwise working fine. There is a known issue in the EM where it will incorrectly display the text of a comment that contains the string "dstaddr" from the repl_cron.cfg file. If you're comfortable at the command line, you can remove this comment line from the file (but make sure you make a backup copy first and be sure not to delete the actual --dstaddr line accidentally). This will be fixed in a future release.

Rob, the issue in the service requests you sent me were somewhat convoluted. I don't want to give too much detail since this is a public forum. Only some of these apply to your particular situation but when dealing with these types of EM issues, the most common problems are:

1. Attempting to monitor a 4.x or 5.x grid from a 6.x Enterprise Manager without first installing the required Administrator Server "web services" SDK. The way EM interfaces with the Administrator Server changed in 6.x. There are a pair of hotfixes (one for 4.x and one for 5.x) that have to be applied to the older grids before they can be monitored by a 6.x EM.
2. Attempting to monitor a newer grid from an older Enterprise Manager.
3. Name resolution failing (or /etc/hosts files not up to date) or giving incorrect information (this tends to affect NAT environments in particular). Note that the grid hosting the EM and the grid being monitored must be able to resolve each other by name.
4. Ports required by the Enterprise Manager / MCS blocked at the firewall.
5. Routing configuration issues.

Any grid running 6.0.1-65 should be upgraded to 6.0.1-66 if possible. There are several important bugfixes in 6.0.1-66 including a fix for an EM performance issue and a critical fix for Data Domain integration.

Unfortunately this is one of those issues that really has to be looked at case-by-case since there are so many factors involved in the correct operation of the Enterprise Manager, especially when monitoring remote grids.

Ian,  Thanks for the response to Rob.  I will take a look at the repl_cron.cfg file and see what I can see!