hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dejan Menges <dejan.men...@gmail.com>
Subject Re: HBase version mismatch
Date Sat, 22 Oct 2011 22:42:00 GMT
So, I figured out what's happening. As hbase-default.xml is not standard in
$HBASE_HOME/conf anymore in 0.90.4, client can't check version in:

<property skipInDoc="true">
    This defaults file was compiled for version @@@VERSION@@@. This variable
is used
    to make sure that a user doesn't have an old version of
hbase-default.xml on the

Luckily, found this in src folder :)

So, I think this is good starting point for my problem resolution, but I
would anyway like to know, if somebody can help me, more about this option
and it's usage.


On Sun, Oct 23, 2011 at 12:31 AM, Dejan Menges <dejan.menges@gmail.com>wrote:

> Hi,
> I have pretty interesting problem.
> We've had application which was running on Tomcat while we were using HBase
> 0.20.3 and before we decided to migrate to HBase 0.90.4
> As we migrated successfully, I started getting strange error trying to run
> this application:
> Error during initalization of HTablePool - NativeException:
> java.lang.RuntimeException: hbase-default.xml file seems to be for and old
> version of HBase (null), this version is 0.90.4:
>  org/apache/hadoop/hbase/HBaseConfiguration.java:66:in
> `checkDefaultsVersion'
> org/apache/hadoop/hbase/HBaseConfiguration.java:89:in `addHbaseResources'
> This machine don't have any version of HBase or Hadoop (locally) and it's
> configuration only points to remote Zookeeper, which successfully does it's
> job, as other application uses new version without any problem.
> I'm trying to figure out how/where client is finding hbase-default.xml and
> how it's checking it's version, as we don't have that file at all, as a
> recommendation to upgrade to new HBase. Does it means that we actually need
> this file in $HBASE_HOME/conf for client to be able to figure out which
> version we are using?
> I'm also checking HBase code for this implementation, and (maybe it's
> because it's late) I'm not able to figure out what's happening.
> Regards,
> Dejan

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message