You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@impala.apache.org by Jim Apple <jb...@cloudera.com> on 2017/07/09 22:52:40 UTC
Can't start minicluster
I am getting the following message in FATAL when I try to start a
minicluster
Check failed: _TImpalaQueryOptions_VALUES_TO_NAMES.size() ==
TImpalaQueryOptions::DEFAULT_JOIN_DISTRIBUTION_MODE + 1 (57 vs. 56)
Any ideas what is going on? I was actually trying to buildall.sh
-format_metastore -format_sentry_policy_db because I was seeing messages
like the following (in hive.log) when I tried to start the minicluster:
org.postgresql.util.PSQLException: ERROR: column A0.SCHEMA_VERSION_V2 does
not exist
Re: Can't start minicluster
Posted by Jim Apple <jb...@cloudera.com>.
Using that command and then running a build without -noclean WFM. Thanks!
On Sun, Jul 9, 2017 at 9:10 PM, Tim Armstrong <ta...@cloudera.com>
wrote:
> Maybe the thrift be/generated-sources are out of sync with the source code?
>
> We had some kind of metastore scheme upgrade that caused the other one.
> Dimitris' instructions to fix them were:
>
> > To fix this without doing a full data reload, you can use the following
> command:
> > ${IMPALA_TOOLCHAIN}/cdh_components/hive-1.1.0-cdh5.13.0-
> SNAPSHOT/bin/schematool
> -upgradeSchema -dbType {type}
> > where type is one of 'postgres' or 'mysql', depending on your setup.
>
> On Sun, Jul 9, 2017 at 3:52 PM, Jim Apple <jb...@cloudera.com> wrote:
>
> > I am getting the following message in FATAL when I try to start a
> > minicluster
> >
> > Check failed: _TImpalaQueryOptions_VALUES_TO_NAMES.size() ==
> > TImpalaQueryOptions::DEFAULT_JOIN_DISTRIBUTION_MODE + 1 (57 vs. 56)
> >
> > Any ideas what is going on? I was actually trying to buildall.sh
> > -format_metastore -format_sentry_policy_db because I was seeing messages
> > like the following (in hive.log) when I tried to start the minicluster:
> >
> > org.postgresql.util.PSQLException: ERROR: column A0.SCHEMA_VERSION_V2
> > does
> > not exist
> >
>
Re: Can't start minicluster
Posted by Tim Armstrong <ta...@cloudera.com>.
Maybe the thrift be/generated-sources are out of sync with the source code?
We had some kind of metastore scheme upgrade that caused the other one.
Dimitris' instructions to fix them were:
> To fix this without doing a full data reload, you can use the following
command:
> ${IMPALA_TOOLCHAIN}/cdh_components/hive-1.1.0-cdh5.13.0-
SNAPSHOT/bin/schematool
-upgradeSchema -dbType {type}
> where type is one of 'postgres' or 'mysql', depending on your setup.
On Sun, Jul 9, 2017 at 3:52 PM, Jim Apple <jb...@cloudera.com> wrote:
> I am getting the following message in FATAL when I try to start a
> minicluster
>
> Check failed: _TImpalaQueryOptions_VALUES_TO_NAMES.size() ==
> TImpalaQueryOptions::DEFAULT_JOIN_DISTRIBUTION_MODE + 1 (57 vs. 56)
>
> Any ideas what is going on? I was actually trying to buildall.sh
> -format_metastore -format_sentry_policy_db because I was seeing messages
> like the following (in hive.log) when I tried to start the minicluster:
>
> org.postgresql.util.PSQLException: ERROR: column A0.SCHEMA_VERSION_V2
> does
> not exist
>