Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
HBase, mail # dev - 0.94.6.1 discussion (WAS [jira] [Commented] (HBASE-8259) Snapshot backport in 0.94.6 breaks rolling restarts)


Copy link to this message
-
Re: 0.94.6.1 discussion (WAS [jira] [Commented] (HBASE-8259) Snapshot backport in 0.94.6 breaks rolling restarts)
lars hofhansl 2013-04-04, 18:12
Pulling 0.94.6 is another option.

________________________________
 From: Stack <[EMAIL PROTECTED]>
To: HBase Dev List <[EMAIL PROTECTED]>
Sent: Thursday, April 4, 2013 11:09 AM
Subject: Re: 0.94.6.1 discussion (WAS [jira] [Commented] (HBASE-8259) Snapshot backport in 0.94.6 breaks rolling restarts)
 
+1 on 0.94.7 rather than 0.94.6.1.

Chatting w/ JD and Aleks, we should pull 0.94.6.  I can add note up on
webpage to explain the hole.

Can I help w/ the 0.94.7 rolling?

St.Ack
On Thu, Apr 4, 2013 at 10:51 AM, Jean-Daniel Cryans <[EMAIL PROTECTED]>wrote:

> My problem with delaying getting this patch out there is that people
> who adopt 0.94.6 are putting themselves in a hole because you also
> can't roll restart out of it.
>
> J-D
>
> On Thu, Apr 4, 2013 at 10:28 AM, Ted Yu <[EMAIL PROTECTED]> wrote:
> > Third 0.94.6 RC was cut on Mar. 14th.
> > It has been almost 3 weeks since them. Meaning we have sizable goodies
> for
> > the next release.
> >
> > I would vote for 0.94.7 release which would come in April anyway.
> >
> >
> > On Thu, Apr 4, 2013 at 10:21 AM, Jean-Daniel Cryans <[EMAIL PROTECTED]
> >wrote:
> >
> >> We might wanna bring this discussion here on dev@.
> >>
> >> So right now the questions are:
> >>
> >>  - Should we call a release with just HBASE-8259 as 0.94.6.1 or
> >> 0.94.7? How much testing are we expecting from folks?
> >>  - Should we just cut a release with what's in the branch and call it
> >> 0.94.7?
> >>
> >> My opinion is to release 0.94.6.1 with just HBASE-8259 to replace the
> >> current 0.94.6. Have the normal unit test run and release based on the
> >> +1s we gather.
> >>
> >> J-D
> >>
> >> On Thu, Apr 4, 2013 at 10:01 AM, Ted Yu (JIRA) <[EMAIL PROTECTED]> wrote:
> >> >
> >> >     [
> >>
> https://issues.apache.org/jira/browse/HBASE-8259?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13622514#comment-13622514
> ]
> >> >
> >> > Ted Yu commented on HBASE-8259:
> >> > -------------------------------
> >> >
> >> > What level of verification effort are you expecting ?
> >> > If normal procedure of validating a release is involved, 0.94.7
> release
> >> seems to be a better fit.
> >> >
> >> >> Snapshot backport in 0.94.6 breaks rolling restarts
> >> >> ---------------------------------------------------
> >> >>
> >> >>                 Key: HBASE-8259
> >> >>                 URL:
> https://issues.apache.org/jira/browse/HBASE-8259
> >> >>             Project: HBase
> >> >>          Issue Type: Bug
> >> >>    Affects Versions: 0.94.6
> >> >>            Reporter: Jean-Daniel Cryans
> >> >>            Assignee: Matteo Bertozzi
> >> >>            Priority: Blocker
> >> >>             Fix For: 0.94.7
> >> >>
> >> >>         Attachments: HBASE-8259-v0.patch
> >> >>
> >> >>
> >> >> [~aleksshulman] found with his nifty QA tools that 0.94.6 has an
> >> incompatible change due to HBASE-7360 (Snapshot 0.94 Backport) that
> breaks
> >> rolling restarts.
> >> >> RegionTransitionData.write() uses eventType.ordinal() that is the
> index
> >> in the enum and not the value specified in the enum definition. It
> means we
> >> can't add new states in the middle of the list. This can be fixed by
> moving
> >> C_M_SNAPSHOT_TABLE and C_M_RESTORE_SNAPSHOT at the end of the list.
> Trunk
> >> does the right thing already.
> >> >> Right now, RIT znodes created with 0.94.6 (or top of the branch) will
> >> use the wrong value for the event type. You will see things like:
> >> >> {noformat}
> >> >> 2013-04-03 14:57:25,197 DEBUG
> >> org.apache.hadoop.hbase.zookeeper.ZKAssign:
> >> regionserver:60020-0x13dd1e10dbd0004 Attempting to transition node
> >> 70236052/-ROOT- from M_ZK_REGION_OFFLINE to RS_ZK_REGION_OPENING
> >> >> 2013-04-03 14:57:25,201 WARN
> >> org.apache.hadoop.hbase.zookeeper.ZKAssign:
> >> regionserver:60020-0x13dd1e10dbd0004 Attempt to transition the
> unassigned
> >> node for 70236052 from M_ZK_REGION_OFFLINE to RS_ZK_REGION_OPENING
> failed,
> >> the node existed but was in the state M_SERVER_SHUTDOWN set by the