Re: [ANN] hbase-0.2.0 Release Candidate 1

[Rick Hangartner]

See below.

On Jul 23, 2008, at 11:25 AM, Jean-Daniel Cryans wrote:
Rick,

When you deleted the hbase directory and it's content, you also  
deleted the
version file and this is why it told you to install 0.1.x first and  
migrate
it.

Ahh, that makes sense.  We thought something like this could be afoot,  
perhaps because version info was stored in one or more places that  
could end up "out of sync" after deleting directories and/or files.   
But we weren't sure why Hbase started up without saying we needed to  
do a migrate after we blew away the Hadoop disk, since in that case  
the disk also wouldn't have had an hbase directory or a version file  
in that directory at all the first time HBase is started after that.

I perhaps should mention we did see a case another issue with the  
version file.   As best I can recall, once when we blew away the  
Hadoop disk and rebuilt it HBase threw an exception because of an  
empty version file when we started up "bin/hbase shell" for the first  
time.  When we looked at the HBase FSUtils.java code and figured out  
what it was looking at and for, indeed the version file existed but  
was empty.  When we just deleted that file and re-started HBase it,  
and the shell, then worked just fine.



So after to first step (2.a) it was still complaining about redo logs?

(2.a) was the one case we didn't repeat from start to finish carefully  
checking each step because of time limitations (I can't be sure we  
didn't accidentally blow away the version file on our first try).  At  
the time, I had a suspicion if we had done repeated that case  
carefully it would have succeeded.

But no, it wasn't complaining about redo logs after that even when we  
tried to re-run just the "bin/hbase migrate upgrade" operation.

Thanks,
Rick



J-D

On Wed, Jul 23, 2008 at 2:11 PM, Rick Hangartner <[EMAIL PROTECTED] 
>
wrote:

Hi Jean-Daniel,

We just tried migrating test data and didn't get that to work.

However, we're at a stage in this project where we are just  
finishing a new
application with new table(s) so we don't have "real" data to  
migrate.  In
the end, then, the "upgrade" for us worked because HBase-0.2.0 ran  
from a
clean install, and so far the pieces of the app I've already ported  
pass the
tests we have just fine on HBase-0.2.0.  In the next several days,  
we'll
know a little more as we re-work other pieces of the app and  
perhaps make
some other attempts at upgrading test data.

(Another group here has an app in production using HBase-0.1.x.  I  
think
they may wait until we get our project up and have some experience  
with
HBase-0.2.x before upgrading.)

Thanks,
Rick



On Jul 23, 2008, at 10:50 AM, Jean-Daniel Cryans wrote:

Rick,

So you are saying that upgrading from 0.1.3 to 0.2.0 didn't work  
for you?

Regards your question, I strongly suggest that you develop for the  
0.2
branch. If you think there is some features that are missing, feel  
free to
fill a Jira and this will solve your problem. Or you can even  
contribute a
patch.

J-D

On Wed, Jul 23, 2008 at 1:40 PM, Rick Hangartner <[EMAIL PROTECTED] 
>
wrote:

Hi,

All in all, we like what we've seen of HBase-0.2.0 so far and the  
API
changes.

Our upgrade test with Release Candidate 1 was a little bumpy but
generally
positive.

We set up a single virtualized test machine (in this case VMWare  
Fusion
because an OS X machine was close at hand, although we also run Xen
virtualized installs), with a virtual disk, running hadoop-0.16.4  
and
hbase-0.1.3 with a small test table in a Debian install so we could
rapidly
test alternative upgrades paths and data migration.

We tried our own upgrade procedure once based on some previous  
experience
and we tried the "formal" upgrade procedure  found at:

http://wiki.apache.org/hadoop/Hbase/HowToMigrate
http://wiki.apache.org/hadoop/Hadoop%20Upgrade

In each case, we of course started with a fresh copy of our  
virtualized
machine.

Here's what we found:

1) In both approaches, the upgrade from hadoop-0.16.4 to  
hadoop-0.17.1
seem
to go OK.

2) In both approaches, when we tried to do the data migration from
hbase-0.1.3 to hbase-0.2.0 we first got migration failures due to
"unrecovered region server logs". Following the 'Redo Logs'  
comments in
the
"http://wiki.apache.org/hadoop/Hbase/HowToMigrate"; doc, starting  
afresh
with a  new copy of our virtualized system each time, we tried  
these
methods
of getting rid of the those logs and the fatal error:

a) deleting just the log files in the "/hbase" directory
b) deleting the entire contents of the "/hbase" directory (which  
means
we
lost our data, but we are just investigating the upgrade path,  
after all)
c) deleting the "/hbase" directory entirely and creating a new "/ 
hbase"
directory.

I should also note that we would need to repeat approach a) to be  
100%
certain of our results for that case.  (We've already repeated  
approaches
b)
and c) and have just run out of time for these upgrade tests  
because we
need
to get to other things).

In all cases, the migrate then failed as:

[EMAIL PROTECTED]:~/hbase$ bin/hbase migrate upgrade
08/07/22 18:03:16 INFO util.Migrate: Verifying that file system is
available...
08/07/22 18:03:16 INFO util.Migrate: Verifying that HBase is not
running...
08/07/22 18:03:17 INFO ipc.Client: Retrying connect to server:
savory1/10.0.0.45:60000. Already tried 1 time(s).
...
08/07/22 18:03:26 INFO ipc.Client: Retrying connect to server:
savory1/10.0.0.45:60000. Already tried 10 time(s).
08/07/22 18:03:27 INFO util.Migrate: Starting upgrade
08/07/22 18:03:27 FATAL util.Migrate: Upgrade failed
java.io.IOException: Install 0.1.x of hbase and run its migration  
first
   at org.apache.hadoop.hbase.util.Migrate.run(Migrate.java:181)
   at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:65)
   at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:79)
   at org.apache.hadoop.hbase.util.Migrate.main(Migrate.java:446)

As you can see, we were then in a bit of a Catch-22.  To re-install
HBase-0.1.x also required re-installing Hadoop-0.16.4 (we tried
reinstalling
HBase-0.1.x without doing that!) so there was no way to proceed.
Attempting
to start up HBase-0.2.0 just resulted in an error message that we  
needed
to
do the migrate.

3) Since this was just a test, we then blew away the disk used by  
Hadoop
and re-built the namenode per a standard Hadoop new install.
Hadoop-0.17.1
and Hbase-0.2.0 then started up just fine.  We only ran a few  
tests with
the
new Hbase command line shell in some ways we used the old HQL  
shell for
sanity checks, and everything seems copacetic.

A few other comments:

- The new shell takes a bit of getting used to, but seems quite
functional
(we're not the biggest Ruby fans, but hey, someone took this on and
upgraded
the shell so we just say: Thanks!)

- We really like how timestamps have become first-class objects  
in the
HBase-0.2.0 API .  Although, we were in the middle of developing  
some
code
under HBase-0.1.3 with workarounds for timestamps not being first- 
class
objects and we will have to decide whether we should backup and
re-develop
for HBase-0.2.0 (we know we should), or plunge ahead with what we  
were
doing
under HBase-0.1.3 just to discard it in the near future because  
of the
other
advantages of HBase-0.2.0.  Is there anything we should consider in
making
this decision, perhaps about timing of any bug fixes and an  
official
release
of HBase-0.2.0? (HBase-0.2.1?).

Thanks for what continues to be a very useful and interesting  
product.

Rick


On Jul 22, 2008, at 3:24 PM, stack wrote:

The first 0.2.0 release candidate is available for download:


http://people.apache.org/~stack/hbase-0.2.0-candidate-1/<http://people.apache.org/%7Estack/hbase-0.2.0-candidate-1/ 
>
<http://people.apache.org/%7Estack/hbase-0.2.0-candidate-1/>

Please take this release candidate for a spin. Check the  
documentation,
that unit tests all complete on your platform, etc.

Should we release this candidate as hbase 0.2.0?  Vote yes or no  
before
Friday, July 25th.

Release 0.2.0 has over 240 issues resolved [1] since the branch  
for 0.1
hbase was made.  Be warned that hbase 0.2.0 is not backward  
compatible
with
the hbase 0.1 API.  See [2] Izaak Rubins' notes on the high- 
level API
differences between 0.1 and 0.2.  For notes on how to migrate  
your 0.1
era
hbase data to 0.2, see Izaak's migration guide [3].

Yours,
The HBase Team

1.

https://issues.apache.org/jira/secure/ReleaseNote.jspa?version=12312955&styleName=Html&projectId=12310753&Create=Create
2. http://wiki.apache.org/hadoop/Hbase/Plan-0.2/APIChanges
3. http://wiki.apache.org/hadoop/Hbase/HowToMigrate