Replication factor being set incorrectly

[GoogleCodeExporter]

When starting up a cluster with less than three nodes voldemort and mysql
still attempt to use 3x replication causing startup to fail.

Original issue reported on code.google.com by jmkupfer...@gmail.com on 31 Mar 2010 at 6:34

[GoogleCodeExporter]

May want to see old code for AppScale 1.3 that did this correctly for guidance.

Original comment by shattere...@gmail.com on 31 Mar 2010 at 6:39

• Added labels: Milestone-Release1.4, Component-Logic, Usability

[GoogleCodeExporter]

The way replication factor was previously being set no longer works as a result 
of
the advanced layout changes. Replication factor will now become a part of the
node_layout code.

Original comment by jmkupfer...@gmail.com on 7 Apr 2010 at 8:55

• Changed state: Started

[GoogleCodeExporter]

Replication factor is now being verified and set in the node_layout code 
committed in
revision 125 of the tools.

Original comment by jmkupfer...@gmail.com on 7 Apr 2010 at 10:34

• Changed state: Fixed

[GoogleCodeExporter]

HBase and Hypertable could work in single node and 2 nodes now.
Should we allow that?

Original comment by yoshi...@gmail.com on 9 Apr 2010 at 2:54

• Changed state: Started

[GoogleCodeExporter]

Certainly. The last submitted paper had results for HBase and Hypertable for 2 
nodes,
so we should try to allow for as many deployments as the underlying DBs support.

Original comment by shattere...@gmail.com on 9 Apr 2010 at 3:31

• Added labels: Priority-Critical
• Removed labels: Priority-Medium

[GoogleCodeExporter]

Commit 127 to the tools now allow hbase and hypertable deployments with one or 
two nodes.

Original comment by jmkupfer...@gmail.com on 9 Apr 2010 at 8:08

[GoogleCodeExporter]

The replication factor is now being set appropriately based on the provided 
command
line input and/or the number of nodes specified.

Original comment by jmkupfer...@gmail.com on 9 Apr 2010 at 9:17

• Changed state: Fixed