Multi-server TSE

[rfpup]

TSE 4.1sp2 with two servers; 1A and 2A, each of two gateways equipped with LSU-E.
Server 1 controls gateway 1 and server 2 controls gateway 2
If server 1 fail can the extensions at gateway 2 work normally (call each-other, call the PSTN through the trunks in this gateway)?

 

[doktor]

No, this will not be possible, unless you are using a redundant third server or HLR (only IP extensions). Recommended is a third redundancy server that will take over if one of the two servers (LIM-servers) fail.

///doktor

 

[rfpup]

Hi doctor,
thank you for your information.
Is it possible that this third server could take over any of the two servers in case any of them could fail?
The two ones are currently in different subnets connected only via one ethernet port in the network.
IP of Server 1 is xx.xx.19.xx/25
IP of Server 2 is xx.xx.24.xx/25

How the third server could be integrated in the solution for such configuration?

 

[himdp]

Either doktor or I have misunderstood the question. So please clarify -
I believe you have two LIMs with trunks in each and you are asking if LIM 1 fails can LIM 2 extension still make calls?

My answer would be yes, without a standby server. Unless I have not understood correctly.

 

[rfpup]

Hi himdp,

You have understand correctly,
I have two lims (2 mgw), each of them connected to a specific HP server through a LSU-E and each of them has its own trunks to PSTN.
I asked if lim 1 fails (if server 1 fail) could the extensions of lim 2 work so calling each other and PSTN through PSTN trunks connected to this remaining LIM, lim2 (mgw2).

2. Is it possible that a standby server could be placed somehow in the network to replace each of the two current HP servers in case any of them may fail?
Thank You in advance for your information

 

[himdp]

The normal servers will have a physical IP address and an alias IP address for each interface. When one normal server fails, the alias addresses are taken by the standby server. For this reason, all servers must be in the same network. You must also have network redundancy.

In your case, if you wanted to go for redundancy, you would need a standby server for each main server and a redundant network for each cluster. Also, this must be configured at install time so If you have an existing system it must be re-installed.

Having said that you did not mention that you needed redundancy until doktor sggested it. I maintain that you will get service in mgw 2A if server 1 fails, so do you really want standby servers?

 

[rfpup]

Hi himdp,
thank you for your information.
I think its not necessary the real standby server, but a ready-to be replaced server in order to replace any of the existing one in case of fail.

Has someone tested the VMware for such purpose?
If yes which could be some suggestions?