Re: Safest DB modification to repeat.

[Love Hörnquist Åstrand <lha@kth.se>]

Dave Love <d.love@dl.ac.uk> writes:

> "Henry B. Hotz" <hotz@jpl.nasa.gov> writes:
>
>> I'm setting up a cron job to generate some iprop traffic on a regular
>> basis.  First thought was "cpw -r iproptraffic", but that cycles the
>> kvno, which would be "bad" at some point.
>
> For what it's worth, frobbing a principal's attributes doesn't change
> the kvno.  Presumably you'd have to cycle an attribute to generate the
> traffic, but that would just mean separate cron entries.
>
> This does sound like a good idea for monitoring problems in the
> propagation.

More modern versions of the iprop protocol, have a keep alive message that
that the slave sends now and then to make sure the server knows its
alive. We used to have problems there wasn't enough updates going over the
connection, so monitoring didn't pick up slaves that died.

Love

PGP signature