It depends on what the replication settings are on a given database. You
can utilize PIRC (purge interval replication control) on the databases in
the cluster member(s) that are still active to prevent ghost documents
from being replicated in when the cluster member that has been down for an
extended period of time is brought back online.
Hope this helps,
Kim
----------------------------------------------------------------------------------------------
Kim Greene
Kim Greene Consulting, Inc.
President
IBM Collaboration Solutions & MongoDB
Office: 507-216-5632 Mobile: 507-254-9353
Email: kim@xxxxxxxxxxxxx Web: www.kimgreene.com
Blog: DominoDiva LinkedIn: kimgreeneconsulting Twitter:
iSeriesDomino
From: "Rob Berendt" <rob@xxxxxxxxx>
To: "domino400 domino400 domino400" <domino400@xxxxxxxxxxxxxxxxxx>
Date: 01/06/2021 01:16 PM
Subject: How long can a cluster member be down before it becomes
dangerous?
Sent by: "Domino400" <domino400-bounces@xxxxxxxxxxxxxxxxxx>
I remember reading somewhere that if a cluster member is down for an
extended time, such as a week or so, that bringing it back online can be
dangerous. Meaning something like this:
ServerA is in a cluster with ServerB.
ServerA is dropped for an extended period of time.
Databases on ServerB get new records, etc.
Since ServerA was down so long if you bring it back up resyncing it may
actually drop the new rows or some such thing. It's all a little foggy to
me exactly what the concern was. Maybe it was deletions in ServerB would
get replicated back in from ServerA. This seems more likely.
I tried a google search but apparently I'm not getting the right
terminology to find a related hit.
Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1
Group Dekko
Dept 1600
Mail to: 7310 Innovation Blvd, Suite 104
Ft. Wayne, IN 46818
Ship to: 7310 Innovation Blvd, Dock 9C
Ft. Wayne, IN 46818
http://www.dekko.com
midrange.com
