You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@phoenix.apache.org by Justin Workman <ju...@gmail.com> on 2016/04/19 19:44:29 UTC
SYSTEM.CATALOG Table Is Not Recreated
All,
I had to drop both the SYSTEM.CATALOG and SYSTEM.SEQUENCE tables from hbase
do to what appeared to be corruption with these two tables.
I expected that they would be recreated when I first reconnected with
sqlline, but they are not. I do see in the Hbase master logs that the
create table statement for SYSTEM.CATALOG should be executed, and I do not
see any errors around the create statement in the logs, however the table
is never created.
Any pointers on how to get these two tables recreated?
Thanks
Justin
Re: SYSTEM.CATALOG Table Is Not Recreated
Posted by Justin Workman <ju...@gmail.com>.
Yes. When restarting the client I get an error about the table not found. And if I leave sqlline just sitting there, I see the create statement appear over and over in the master logs.
Sent from my iPhone
> On Apr 19, 2016, at 2:29 PM, James Taylor <ja...@apache.org> wrote:
>
> I'm assuming you restarted your client too (i.e. the client JVM)?
>
>> On Tuesday, April 19, 2016, Justin Workman <ju...@gmail.com> wrote:
>> I have restarted the cluster several times, with no luck. Any other ideas? HBase is up, but I cannot connect via sqlline because this table does not exist.
>>
>>> On Tue, Apr 19, 2016 at 12:28 PM, James Taylor <ja...@apache.org> wrote:
>>> Make sure you bounce your cluster too (or at a minimun the region server hosting the SYSTEM.CATALOG), as you must be picking up the server cached table.
>>>
>>>
>>>> On Tuesday, April 19, 2016, Justin Workman <ju...@gmail.com> wrote:
>>>> All,
>>>>
>>>> I had to drop both the SYSTEM.CATALOG and SYSTEM.SEQUENCE tables from hbase do to what appeared to be corruption with these two tables.
>>>>
>>>> I expected that they would be recreated when I first reconnected with sqlline, but they are not. I do see in the Hbase master logs that the create table statement for SYSTEM.CATALOG should be executed, and I do not see any errors around the create statement in the logs, however the table is never created.
>>>>
>>>> Any pointers on how to get these two tables recreated?
>>>>
>>>> Thanks
>>>> Justin
Re: SYSTEM.CATALOG Table Is Not Recreated
Posted by James Taylor <ja...@apache.org>.
I'm assuming you restarted your client too (i.e. the client JVM)?
On Tuesday, April 19, 2016, Justin Workman <ju...@gmail.com> wrote:
> I have restarted the cluster several times, with no luck. Any other ideas?
> HBase is up, but I cannot connect via sqlline because this table does not
> exist.
>
> On Tue, Apr 19, 2016 at 12:28 PM, James Taylor <jamestaylor@apache.org
> <javascript:_e(%7B%7D,'cvml','jamestaylor@apache.org');>> wrote:
>
>> Make sure you bounce your cluster too (or at a minimun the region server
>> hosting the SYSTEM.CATALOG), as you must be picking up the server cached
>> table.
>>
>>
>> On Tuesday, April 19, 2016, Justin Workman <justinjworkman@gmail.com
>> <javascript:_e(%7B%7D,'cvml','justinjworkman@gmail.com');>> wrote:
>>
>>> All,
>>>
>>> I had to drop both the SYSTEM.CATALOG and SYSTEM.SEQUENCE tables from
>>> hbase do to what appeared to be corruption with these two tables.
>>>
>>> I expected that they would be recreated when I first reconnected with
>>> sqlline, but they are not. I do see in the Hbase master logs that the
>>> create table statement for SYSTEM.CATALOG should be executed, and I do not
>>> see any errors around the create statement in the logs, however the table
>>> is never created.
>>>
>>> Any pointers on how to get these two tables recreated?
>>>
>>> Thanks
>>> Justin
>>>
>>
>
Re: SYSTEM.CATALOG Table Is Not Recreated
Posted by Justin Workman <ju...@gmail.com>.
I have restarted the cluster several times, with no luck. Any other ideas?
HBase is up, but I cannot connect via sqlline because this table does not
exist.
On Tue, Apr 19, 2016 at 12:28 PM, James Taylor <ja...@apache.org>
wrote:
> Make sure you bounce your cluster too (or at a minimun the region server
> hosting the SYSTEM.CATALOG), as you must be picking up the server cached
> table.
>
>
> On Tuesday, April 19, 2016, Justin Workman <ju...@gmail.com>
> wrote:
>
>> All,
>>
>> I had to drop both the SYSTEM.CATALOG and SYSTEM.SEQUENCE tables from
>> hbase do to what appeared to be corruption with these two tables.
>>
>> I expected that they would be recreated when I first reconnected with
>> sqlline, but they are not. I do see in the Hbase master logs that the
>> create table statement for SYSTEM.CATALOG should be executed, and I do not
>> see any errors around the create statement in the logs, however the table
>> is never created.
>>
>> Any pointers on how to get these two tables recreated?
>>
>> Thanks
>> Justin
>>
>
Re: SYSTEM.CATALOG Table Is Not Recreated
Posted by James Taylor <ja...@apache.org>.
Make sure you bounce your cluster too (or at a minimun the region server
hosting the SYSTEM.CATALOG), as you must be picking up the server cached
table.
On Tuesday, April 19, 2016, Justin Workman <ju...@gmail.com> wrote:
> All,
>
> I had to drop both the SYSTEM.CATALOG and SYSTEM.SEQUENCE tables from
> hbase do to what appeared to be corruption with these two tables.
>
> I expected that they would be recreated when I first reconnected with
> sqlline, but they are not. I do see in the Hbase master logs that the
> create table statement for SYSTEM.CATALOG should be executed, and I do not
> see any errors around the create statement in the logs, however the table
> is never created.
>
> Any pointers on how to get these two tables recreated?
>
> Thanks
> Justin
>