Help with replication ..

First, let me tell you what we are trying to accomplish. We are looking to setup 2 ERAS and have one as the 'default' and one as a 'failover' so when the default goes down, clients have a fallback.

I setup the second unit, let's call them BOB and SUE. BOB is the primary that all clients currently connect to. SUE is the new kid on the block.

I setup BOB to 'replicate from' SUE, and SUE to 'replicate to' BOB. I set the default policy all clients are under to use both servers (IE, primary and secondary).

Now, the problems. If I check SUE and do a "REPLICATE UP NOW", nothing on SUE changes. No clients show up. No policies or groups appear. Nothing. If I check BOB, I don't know how to check to see if the replication came down. I didn't see anything in the logs.

Checking my clients, the Mac client didn't get the setting for the secondary server. Checking the Windows clients and they did get the setting.

So, what am I doing wrong here? I tried to set the 'replicate to' on BOB to SUE and it yelled at me since it was allowing the 'replicate from' from SUE.

I was thinking of trying IP's ... using SUE's IP in the 'replicate to' on BOB and using BOB's IP in the 'replicate from' on SUE. Is there a better way of doing this?

EDIT: Since this is a test setup w/10 seats I gave it a go ...

If I set the 'replicate to' setting to the IP of the other server, and the 'replicate from' setting to the host name of the other server, I was able to get it to transfer policy and clients, and clients are talking to it now; but groups didn't come over. Also, I cannot edit the policy assigned to the clients. Setup as below:

BOB = 192.168.10.35
SUE = 192.168.10.24

BOB - replicate to 192.168.10.24 / replicate from SUE
SUE - replicate to 192.168.10.35 / replicate from BOB

EDIT 2: I guess clients aren't talking to it ... it appears the timers are only copied from the last update; IE, when I pushed a replication from BOB -> SUE and SUE -> BOB it grabbed the timers from the last time the clients talked to BOB, and is now counting from those....

 

Hello,

Clients will only attempt to connect to the secondary server for remote administrator if the first server is unreachable. When you're testing, you will either have to power off BOB, or unplug the network cable for it.

As far as logging is concerned, when you're testing replication you should change the verbosity to level 5+ by going to Tools > Server options > Logging tab and change the log verbosity.

Here is a replication article for your reference, if you haven't seen it yet. One caveat that hooks many users is you should also go to the security tab and either set a password for replication, or check the box below the passwords that is labeled "Enable unauthenticated access for Replication".

If you require further assistance, I encourage you to submit a support case with Eset Customer Care.

Thanks,

Matt

 

Does replication copy down policies and groups? or do those have to be manually updated?

I guess I am a little confused. So, would the backup server be the one I am replicating from? IE, it replicates up to the master, and the master has all the clients; and when the master goes down, they fail over to the backup since it would be setup as the secondary?

Essentially here's what I want: 2 identical servers, 1 setup as a primary and 1 as a secondary, in the event the primary goes down, the clicks will fallback to the secondary for both updates and administration. Can you give me an idea of how to do this or point me at an article that explains it? I was looking at clustering but I am lost on that one.

 

Ok, I got it figured out I think... please chime in if I am wrong.

So, using the BOB/SUE examples above ...
BOB is the primary server clients connect to, SUE is the new failover serever

In BOB, I set it to allow replication from SUE, I configure the remote admin in the policy to point to BOB as the primary and SUE as the secondary; updates from BOB.

In SUE, I set it to replicate up to BOB and configure the policy the same as BOB's except updates will come from SUE now. I set a policy rule to force the policy change if the client connecting is coming from a different primary server.

Testing shows the following happens ---
1. BOB is up and running, clients connect and update fine.
2. BOB goes down, clients start connecting to SUE and picking up SUE's policy (which makes them check for BOB first and failing him use SUE)
3. BOB comes back up, clients begin using BOB again.