slave index much larger than master index

[anna headley <an...@gmail.com>]

Hello solr-user list,

I appear to have a number of issues right now on my slave server.

1.  The most confusing one is that my slave index is currently 67 gigs, but
my master index is only 27 gigs.  Has anyone seen this before?  Has anyone
an idea of what could cause this?

2.  I haven't been able to find good information on the
replication.properties file.  I have set the master to replicate on
optimize, which happens once per day.  On the slave server, replication is
happening successfully once per day but my timesFailed counter goes up and
up.  The pollInterval has been set to 00:00:20 (I am planning to change
this since it is really too frequent).  Does Solr consider it a replication
failure if the slave polls but the master has not set the signal to
replicate?

3.  How do I interpret the numbers in replication.properties under
replicationFailedAt and indexReplicatedAt, etc.?

Thanks for any help you can offer.

Anna