2009/9/29 Lawrence Giam <[email protected]>

>  Hi All,
>
>
>
> I am testing failover with the following setup:
>
> N1 - Master
>
> N2 – Slave
>
>
>
> To test failover, I issue this command: slonik_failover 1 2 | slonik
>
> -------------------------------------------
>
> INFO: calling failedNode(1,2) on node 1
>
> <stdin>:4: NOTICE:  failedNode: set 1 has no other direct receivers - move
> now
>
> INFO: Waiting for slon engines to restart
>
> INFO: Node with highest sync for set 1 is 2
>
> <stdin>:10: Replication sets originating on 1 failed over to 2
>
> -------------------------------------------
>
>
>
> Now I try to remove the N1 from the setup with this command:
> slonik_drop_node 1 | slonik
>
> <stdin>:4: Error: Node ID and event node cannot be identical
>
>
>
> I also try this : slonik_drop_node 2 | slonik
>
> <stdin>:10: dropped node 2 cluster
>
>
>

yikes ;-)


> Problem is when I check the slon log on node2, it is reporting:
>
> ----------------------
>
> 2009-09-29 13:49:19 SGT DEBUG1 slon: restart of worker
>
> 2009-09-29 13:49:19 SGT CONFIG main: slon version 1.2.16 starting up
>
> 2009-09-29 13:49:19 SGT DEBUG2 slon: watchdog process started
>
> 2009-09-29 13:49:19 SGT DEBUG2 slon: watchdog ready - pid = 28749
>
> 2009-09-29 13:49:19 SGT DEBUG2 slon: worker process created - pid = 29314
>
> 2009-09-29 13:49:19 SGT ERROR  cannot get sl_local_node_id - ERROR:
> schema "_testrepl" does not exist
>
> 2009-09-29 13:49:19 SGT FATAL  main: Node is not initialized properly -
> sleep 10s
>
> 2009-09-29 13:49:29 SGT DEBUG2 slon_retry() from pid=29314
>
> 2009-09-29 13:49:29 SGT DEBUG1 slon: retry requested
>
> 2009-09-29 13:49:29 SGT DEBUG2 slon: notify worker process to shutdown
>
> 2009-09-29 13:49:29 SGT DEBUG2 slon: child terminated status: 0; pid:
> 29314, current worker pid: 29314
>
> ----------------------
>
>
>
> Also checking in PgAdmin, it reports the cluster database on node2 was
> removed. Did I issue the wrong command? How do I remove node1 from the
> cluster after handing the origin to node2 with the failover command?
>
>
>

simple - you have dropped N2 (see "yikes" above) , so you have effectively
trashed the cluster.


-- 
Filip Rembiałkowski
JID,mailto:[email protected]
http://filip.rembialkowski.net/
_______________________________________________
Slony1-general mailing list
[email protected]
http://lists.slony.info/mailman/listinfo/slony1-general

Reply via email to