What's the recommendation for the best way for detecting replication
slave replication health, using Solr 1.4 Java replication?  The
current "details" command returns a lot of useful data, but it's not
obvious how best to extract the answer to the simple question: "is
replication ok, slow or failed?"

For example,  it would be helpful if there were a pair of status
values that summarized the slave information like so:

str name "slavestatus"
OK - last replication poll succeeded or first poll hasn't completed
ERROR - last poll failed
DISABLED - disabled

long name "secondsBehind"
Estimate of replication lag.  Maybe min(seconds diff between master
and slave indexVersion, seconds since last successful poll).  No value
reported if the master is unreachable.

Shawn

Reply via email to