hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sharma, Avani" <agsha...@ebay.com>
Subject RE: HBase table lost on upgrade
Date Tue, 31 Aug 2010 19:14:54 GMT
Thanks, Stack. Well, I was able to get the basic hbase cluster to run, but now that I am trying
to boost read performance, I am running into stuff that is either not working or I cannot
easily find solutions to on the net.


-----Original Message-----
From: saint.ack@gmail.com [mailto:saint.ack@gmail.com] On Behalf Of Stack
Sent: Tuesday, August 31, 2010 11:52 AM
To: user@hbase.apache.org
Subject: Re: HBase table lost on upgrade

On Tue, Aug 31, 2010 at 11:36 AM, Sharma, Avani <agsharma@ebay.com> wrote:
> I just upgraded my 3-node cluster for 0.20.3 to 0.20.6. After I did this, I notice that
some of my tables are missing. I recovered the lost tables using
> add_table.rb.
> But I am starting to have doubts about the stability of Hbase now since I have had to
recover lost tables a number of times now. What is the possible reason that the tables could
have gone missing?

HBase can lose data if you you run with its defaults which put data
into /tmp -- and then your machines are rebooted.  HBase can lose data
if you crash out your regionservers or kill -9 them -- use one of the
0.89 series releases and hadoop 0.20-append if you want hbase to
preserve data across regionserver crashes.

You seem to be having a bit of a tough time coming up to speed on
hbase going by your recent mails to the list.  HBase is not the
easiest system to learn.  It takes a bit of effort (We're working on
making it easier to get going but are not there yet).  If you want to
take the time to keep track of what you are doing so you can replay it
here on the list, we'll help you out.  Do take the time to go over the
'getting started' documentation first though.


View raw message