Leader SigSvr down Follower no NRS?

[SL1M1]

RLS 5.5 CS1KM MG
Have recently had an issue with my Leader HP COTS SigSvr 'Internal Fans burned out' and my 'Follower' didn't take over? only have SIP Trunks to a MG1000B and No IP Phones. Looks as if the issue is no NRS component loaded on the 'Follower' as I keep getting an error message indicating this when trying to LogInto NRS, any tips on the easiest process to get it onto the 'Follower' or is it a reload have saved an external back up of the leader NRS

Have Leader SigSvr replaced and swapped out the hard drives SIP Trunks now back up but obviously no redundancy/fail safe still ?

Cheers!!

 

[bignose21]

Pop the Hard Drive out your Leader and stick it in your follower and get the leader back up and running that way.

NRS can only run on Lead Sig Server and you should have an Alternate NRS maybe at your 1000B if no other Nodes on the Network.

 

[SL1M1]

Hi bignose21, OK that'll probably be the reason why I can't login to NRS on the follower at the Main site then.

Those are are set as different Nodes on the Main and 1000B

Do I need to check something is not right on the 1000B or something not ticked within NRS components to handle calls on a Leader Failure ?

As you can see not too knowledgable on this front but thanks for your guidance here

Cheers!!

 

[allenmac]

Like BigNose said, NRS can only run on a Leader, so if your only leader is down, you got no NRS.

If you have 2 Leaders, then designate the 2nd as alternate NRS. More than 2 and the rest get failsafe NRS.

If you only have the 1 node, and you need NRS redundancy, then you may want to look at putting your Primary NRS on a standalone sig server.

 

[SL1M1]

OK Guys so from what you saying if, as was the case, the leader failed due to fan/power failure there is no NRS to route calls over the SIP Trunks unless designate the second SigSvr as 'Alternate NRS' ?

Can I ask your definition then of 'Follower' SigSvr as I've lost the plot

Cheers!!

 

[allenmac]

Each VoIP Node contains 1 or more signaling servers.

There is always a leader, and if you have more than 1 than the others are followers. This is configured when you install the software on the signaling server.

NRS is an application that can be run in 2 configurations: Co-Resident and Standalone.

In Co-Resident mode, there are other applications running on the signaling server, like LTPS, IP Line, Personal Directory, etc.

In Standalone mode, no other applications run on the signaling server, and the NRS role is usually Primary or Alternate.

In my particular environment, My Primary NRS runs on a standalone signaling server. My Alternate NRS runs on the leader signaling server at my headquarters, and all other remote offices run NRS on their leader signaling server in Failsafe mode.

 

[bignose21]

In you enviroment the Ideal Setup sounds like the Leader of your Main site is the Primary NRS and the Leader of your Branch Site is the Alternate.

To do this you need to Tick the enable gatekeeper box on the leader at your branch and set it to alternate. You then define this as the alternate in SIP/H323 GW settings in the node. All the above requires a reboot of the SS.

Log in to the NRS at your Branch and set it up so it knows it is the Alternate and th IP of the Primary.

In the Node at the Main site add the Alternate NRS (Branch Leader TLAN) information to the SIP/H323 GW's (might require a reboot, will prompt).

Now you Either need to Fix your Leader so that the NRS database is synced accross or you can start adding a database to the NRS at the Branch to get you endpoints registered and start passing calls.

Like I said earlier if the Leader and follower are both HP COTS on the Main Site switch them off release the Hard Drives put the one from the Leader in the Follower and Boot it up. Would still suggest putting in the Alternate at the branch just in case you have a Lead SS failure in the future.

 

[SL1M1]

Hi bignose21 and allenmac many thanks for you explanations and suggestions really appreciate it Guys

Already swapped over the hard drives from to get SIP service back up and running whilst we waiting for a replacement COTS HP Server now in place and need to check all those settings highlighted

I did hear that there was a tech bulletin warning about some suspect batches of these HP COTS though

Cheers!!

 

[allenmac]

That tech bulletin is old, and pertained to a specific batch of serial numbered servers. If you had one of those, it was impossible to load the sig server application, as the supplied drivers for several of the server components were not compatible with the nortel installation software.

 

[SL1M1]

OK allenmac so guess its unlucky on that Power/FAN Unit HP SigSvr's only been installed for 10 months

Cheers!!

 

[allenmac]

Yeah. I also had a power supply in one of the HP COTS sig servers die after less than a year in service, and it was one with a serial number after the ones affected by the bulletin.

I have another 7 or 8 that have been in service for almost 2 years now with no issues.

 

[cmoegie]

If you have a leader and follower signaling server on the same node they would both have unique host names but they should both have the same H32ID and SIP ID. That way if the leader fails and the follower takes over it still registered to the NRS with the same SIP ID and it assumes control of the virtual trunks. The only time the alternate NRS should take over is if you have a failure of both the leader and follower on the primary node of the primary NRS.

 

[allenmac]

Almost correct,

NRS does not run on follower signaling servers at all, so it does not matter what state the follower is in at any given time in a node where the leader is also the primary NRS.

It is true that every element in the node should have a unique host name, but share a common H.323 and/or SIP ID.

 

[cmoegie]

NRS does not run on the follower signaling server however if the leader fails the follower becomes the leader and will assume the NRS role and will remain registered with the NRS with the same H323 or SIP ID.
That is the reason for a follower signling server.

 

[bignose21]

Not quite the Follower cannot run NRS. But on the failure of the Leader it is capable of running H323 and SIP Gateway applications (Drop down box against SS in node setup) which will stay registered with the NRS and the Virtual trunk will establish between the Call Server and the H323/SIP Gateways on the Follower.

 

[allenmac]

cmoegie,

I think in this instance you are mistaking the virtual trunk application for NRS.

The virtual trunk application runs on the node master, which is usually the leader signaling server. If the leader sig server fails, then an election is held and if there is a follower signaling server, then it is elected node master, and runs virtual trunk.

If you don't have a follower signaling server, but do have MC-32 cards running LTPS, then they can also become the node master, but they do not run virtual trunk.

So the only way to keep your SIP or H.323 trunks alive if the leader signaling server fails is to have at least 1 follower signaling server.