On Tue, 23 Jan 2007, Jeff Frost wrote:

> I just update a timestamp in a slony_test table every 5 minutes on the master
> and check that on the slaves to make sure it's not more than x minutes old.
> The scripts are simple shell that calls psql.
>
> On Tue, 23 Jan 2007, Mark Stosberg wrote:
>
>> Hello,
>>
>> I'm wondering what people are using to monitor Slony with Nagios. The
>> docs reference:


Start by asking what are the metrics for health:

1) Synchronous number of rows in a each table/view on Masters v.s. Slaves?

Unfortunately, this type check requires that you do more than select 
count(*) from $table_name; Snap/Dump from your Slave/Master.

2) Queue sizes on masters (w/ thresholds)

3) Date/Time of last heartbeat from other nodes

~BAS