Koichi Suzuki - 2015-04-30

We need some amount of work to make BDR available in XC/XL. In
XC/XL, updates to a coordinator is actually handled by (possibly) more
than one datanode. We need global filtering to have logical log
which make sense from applications.

We need 9.5 for BDR, which needs much more work to merge.

In terms of HA, current XC/XL uses log shipping slaves for each
coordinator/datanodes.

Regards;

Koichi Suzuki

2015-04-30 2:18 GMT+09:00 Todd Sykes tsykes04@users.sf.net:

[tickets:#58] Postgres-XL with BDR for data nodes?

Status: open
Milestone: Undecided
Created: Wed Apr 29, 2015 05:18 PM UTC by Todd Sykes
Last Updated: Wed Apr 29, 2015 05:18 PM UTC
Owner: nobody

I am very interested in demonstrating Postgres-XL to my organization to give
us the scaling that we need without going to a NoSQL solution. But, for our
current use case, high availability is massively important. Other solutions
I'm looking at automatically copy shards and handle missing nodes. I've seen
the recommendation to slave each of the nodes and use corosync/pacemaker to
handle HA. Recently, we've been having some success with BDR behind a VIP
for replicated multi-master. Would this be a workable solution in
conjunction with Postgres-XL for the data nodes? How different are the data
nodes than vanilla Postgres?

Sent from sourceforge.net because you indicated interest in
https://sourceforge.net/p/postgres-xl/tickets/58/

To unsubscribe from further messages, please visit
https://sourceforge.net/auth/subscriptions/

 

Related

Tickets: #58