bq: shouldn't the two replicas have the same number of deletions
Not necessarily. We're back to the fact that commits on the replicas in
a single shard fire at different wall clock times. Plus, when segments
are merged, the deleted docs are purged. So it's quite common that
two replicas in the
Something I hadn't know until now. The source cdcr collection has 2 shards
with 1 replica, our target cloud has 2 shards with 2 replicas
Both Source and Target have indexes that are not current
Also we have set all of our collections to ignore external commits
On Thu, Dec 15, 2016 at 1:31 PM,
Looking through our replicas I noticed that in one of our shards (each
shard has 2 replicas)
1 replica shows:
"replicas": [
{
"name": "core_node1",
"core": "sial-catalog-material_shard2_replica2",
"baseUrl": "http://ae1b-ecom-msc04:8983/solr;,
"nodeName": "ae1b-ecom-msc04:8983_solr",
"state":
I am trying to find the reported inconsistencies now.
The timestamp I have was created by our ETL process, which may not be in
exactly the same order as the indexing occurred
When I tried to sort the results by _docid_ desc, solr through a 500 error:
{ "responseHeader":{ "zkConnected":true,
Let's back up a bit. You say "This seems to cause two replicas to
return different hits depending upon which one is queried."
OK, _how_ are they different? I've been assuming different numbers of
hits. If you're getting the same number of hits but different document
ordering, that's a completely
Thanks for the quick feedback.
We are not doing continuous indexing, we do a complete load once a week and
then have a daily partial load for any documents that have changed since
the load. These partial loads take only a few minutes every morning.
The problem is we see this discrepancy long
The commit points on different replicas will trip at different wall
clock times so the leader and replica may return slightly different
results depending on whether doc X was included in the commit on one
replica but not on the second. After the _next_ commit interval (2
seconds in your case), doc
