Ah, I see. I'm not sure if you'd ever see "partial" state since a Phoenix table is represented by multiple rows in the SYSTEM.CATALOG table. Probably not a good idea for SYSTEM.SEQUENCE table as you wouldn't want to see an "old" row (which might make you get duplicate sequence values).
On Thu, Sep 8, 2016 at 1:20 PM, Michael McAllister <[email protected] > wrote: > James > > > > I’m not talking about replication between different clusters. Instead I’m > talking about region replication within the same cluster for High > Availability purposes. An overview is here:- > > > > https://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.3.2/ > bk_hadoop-ha/content/ha-hbase-intro.html > > > > Michael McAllister > > Staff Data Warehouse Engineer | Decision Systems > > [email protected] | C: 512.423.7447 | skype: michael.mcallister.ha > <[email protected]> | webex: https://h.a/mikewebex > > This electronic communication (including any attachment) is confidential. > If you are not an intended recipient of this communication, please be > advised that any disclosure, dissemination, distribution, copying or other > use of this communication or any attachment is strictly prohibited. If you > have received this communication in error, please notify the sender > immediately by reply e-mail and promptly destroy all electronic and printed > copies of this communication and any attachment. > > > > *From: *James Taylor <[email protected]> > *Reply-To: *"[email protected]" <[email protected]> > *Date: *Thursday, September 8, 2016 at 3:04 PM > *To: *user <[email protected]> > > *Subject: *Re: Enabling region replication on Phoenix metadata tables > > > > Take a look at this[1] thread for a discussion on replication of system > tables. You can replicate the SYSTEM.CATALOG table, but you have to be very > careful. Make sure to disable and discard replicated data for > SYSTEM.CATALOG while any Phoenix upgrade is in progress (i.e. first > connection after upgrading to a new minor release) or any physical tables > are being created through DDL statements (which presumably would be run on > both primary and secondary cluster). Otherwise, you'll run into race > conditions between the upgrade code and replication occurring. > > > > > > [1] http://search-hadoop.com/m/9UY0h2JDL91232UBQ1&subj=Replication+ > > > > On Thu, Sep 8, 2016 at 12:03 PM, Michael McAllister < > [email protected]> wrote: > > You did alter specific tables to enable region replication on them, or > create new tables with region replication, correct? > > > > Michael McAllister > > Staff Data Warehouse Engineer | Decision Systems > > [email protected] | C: 512.423.7447 | skype: michael.mcallister.ha > <[email protected]> | webex: https://h.a/mikewebex > > This electronic communication (including any attachment) is confidential. > If you are not an intended recipient of this communication, please be > advised that any disclosure, dissemination, distribution, copying or other > use of this communication or any attachment is strictly prohibited. If you > have received this communication in error, please notify the sender > immediately by reply e-mail and promptly destroy all electronic and printed > copies of this communication and any attachment. > > > > *From: *"Vommi, Nithin" <[email protected]> > *Reply-To: *"[email protected]" <[email protected]> > *Date: *Thursday, September 8, 2016 at 1:47 PM > > > *To: *"[email protected]" <[email protected]> > *Subject: *RE: Enabling region replication on Phoenix metadata tables > > > > I meant, The region servers were never active and did not show up on UI > > > > *From:* Michael McAllister [mailto:[email protected]] > *Sent:* Thursday, September 8, 2016 11:46 AM > *To:* [email protected] > *Subject:* Re: Enabling region replication on Phoenix metadata tables > > > > Nithin > > > > > > > when I tried enabling region replication, I was not able to *bring the > HBase cluster*. > > > > > > > I’m not sure what you mean here. Specifically referring to “bring the > HBase cluster”. > > > > Michael McAllister > > Staff Data Warehouse Engineer | Decision Systems > > [email protected] | C: 512.423.7447 | skype: michael.mcallister.ha > <[email protected]> | webex: https://h.a/mikewebex > > This electronic communication (including any attachment) is confidential. > If you are not an intended recipient of this communication, please be > advised that any disclosure, dissemination, distribution, copying or other > use of this communication or any attachment is strictly prohibited. If you > have received this communication in error, please notify the sender > immediately by reply e-mail and promptly destroy all electronic and printed > copies of this communication and any attachment. > > > > *From: *"Vommi, Nithin" <[email protected]> > *Reply-To: *"[email protected]" <[email protected]> > *Date: *Thursday, September 8, 2016 at 1:42 PM > *To: *"[email protected]" <[email protected]> > *Subject: *RE: Enabling region replication on Phoenix metadata tables > > > > I agree with Michael on enabling region replication for the SYSTEM.CATALOG > table. But when I tried enabling region replication, I was not able to > bring the HBase cluster. I am using HBase 1.2.1 and Phoenix 4.7 (on Amazon > EMR Platform) > > > > Regards > > Nithin > > > > *From:* Michael McAllister [mailto:[email protected] > <[email protected]>] > *Sent:* Thursday, September 8, 2016 11:39 AM > *To:* [email protected] > *Subject:* Enabling region replication on Phoenix metadata tables > > > > Is there any reason why it would be a bad idea to enable region > replication on the Phoenix metadata tables. Specifically, SYSTEM.CATALOG et > al. > > > > From everything I’m reading it seems like it would be a good idea. Those > tables are a single point of failure for Phoenix. If they aren’t up then no > Phoenix connection can be made. > > > > Michael McAllister > > Staff Data Warehouse Engineer | Decision Systems > > [email protected] | C: 512.423.7447 | skype: michael.mcallister.ha > <[email protected]> | webex: https://h.a/mikewebex > > This electronic communication (including any attachment) is confidential. > If you are not an intended recipient of this communication, please be > advised that any disclosure, dissemination, distribution, copying or other > use of this communication or any attachment is strictly prohibited. If you > have received this communication in error, please notify the sender > immediately by reply e-mail and promptly destroy all electronic and printed > copies of this communication and any attachment. > > >
