Enqueue server failover

[Former Member]

Hi,

I just read the presentation about HA-setups for NW04 in SP15/16 ... can anyone give me a hint for the failover mechanism for the enqueue server (on WebAS Java) with the enquere replication server: How long does it take to restart the enquere server in the replication server with

- creation of the initial enqueue tables

- attaching the enqueue server to the replica table

- filling the table created above

- deleting the replication table

and make it running as the new enqueue server?

Can this be done within some seconds?

Thanks for any answers.

Martin

[Former Member · Accepted solution]

Hi!

I'm also interested in this topic. I plan to install a Portal 6.0 in HA using ERS under Unix.

I believe under Windows, things are more straightforward.

Cheers!

[Former Member · Accepted solution]

hi,

Does Enqueue server copy the replication table to newly created table or it just attaches it self to the table owned by replication server and kills the replication process? (Note that Enqueue and Replication tables will be created with same Instance ID in shared memory)

Thanks & Regards,

Venkat

[Former Member · Accepted solution]

Hi Venkat,

If I understand your question correctly...

In an enqueue failover with enqueue replication, the new enqueue process will attach itself to the shared memory of the replication, read in the information, create its own lock table and then kill the replication process. A new replication process is then started up (on another host) so the failover protection can be in place again.

Best Regards,

Matt