Mailing-List: contact user-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hbase.apache.org
Received-SPF: pass (nike.apache.org: domain of harsh@cloudera.com designates
 209.85.210.169 as permitted sender)
Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Apple Message framework v1084)
Subject: Re: Not able to remove entries from .META. table
From: Harsh J <harsh@cloudera.com>
In-Reply-To: 
 <7D9AF4B98807C54EBADEA75DF6D5ACB77FF40263@NDA-HCLC-EVS04.HCLC.CORP.HCL.IN>
Date: Tue, 18 Oct 2011 10:39:38 +0530
Content-Transfer-Encoding: quoted-printable
Message-Id: <F1952295-0568-4EE0-B72B-7F2DCC3BCD24@cloudera.com>
References: 
 <7D9AF4B98807C54EBADEA75DF6D5ACB77FE6F48C@NDA-HCLC-EVS04.HCLC.CORP.HCL.IN>
 <A9C9A3DC-5A3C-4175-9715-1ED87CD12D0F@cloudera.com>
 <CAEf6Z5JV9p_AApmTtwA4Sr0kOhkLGeYEYdd1wC0o1cTVVV8W+A@mail.gmail.com>
 <7D9AF4B98807C54EBADEA75DF6D5ACB77FE6F4AC@NDA-HCLC-EVS04.HCLC.CORP.HCL.IN>
 <CAAha9a1kTx4P=mMjAhoYohb1k6oyZ2bmwo1zkrTwKrNHGGuhpQ@mail.gmail.com>
 <7D9AF4B98807C54EBADEA75DF6D5ACB77FF40263@NDA-HCLC-EVS04.HCLC.CORP.HCL.IN>
To: user@hbase.apache.org

Stuti,

Sorry for the late reply, but Jon's second comment about deletall usage =
is right. It would not work with a specific column "info:regioninfo", =
but rather with just family name "info".

On 18-Oct-2011, at 9:45 AM, Stuti Awasthi wrote:

> Hi Jon,
>=20
> Thanks for the pointers. I figured out that in my case bad table =
created some missing and under replicated blocks in HDFS which was =
causing the problem. Once I removed those under replicated blocks from =
HDFS and retried to delete the entries from .META. table , I was =
successful.
>=20
> Stuti
>=20
> -----Original Message-----
> From: Jonathan Hsieh [mailto:jon@cloudera.com]=20
> Sent: Monday, October 17, 2011 11:06 PM
> To: user@hbase.apache.org
> Subject: Re: Not able to remove entries from .META. table
>=20
> Stuti,
>=20
> I've in the process of completing a offline meta rebuilder.  An early =
version is here: https://issues.apache.org/jira/browse/HBASE-4377
>=20
> It assumes that your file system data is good, so using it would =
require "sidelining" or deleting the regions from the bad table.
>=20
> Jon.
>=20
>=20
> On Fri, Oct 14, 2011 at 5:42 AM, Stuti Awasthi <stutiawasthi@hcl.com> =
wrote:
>=20
>> Hi Harsh,
>>=20
>> I check the entries by the 'get' command :
>>=20
>> hbase(main):001:0> get
>> =
'.META.','system_mappings,,1318585995582.bded23240ed1a28c2b80ece0cbfb73b8.=
','info'
>> COLUMN                             CELL
>> info:regioninfo                   timestamp=3D1318606241130, =
value=3DREGION =3D>
>> {NAME =3D> 'system_mappings,,1318585995582.bded23240ed1a28c2
>>                                       b80ece0cbfb73b8.', STARTKEY =3D>=20=

>> '', ENDKEY =3D> '', ENCODED =3D> bded23240ed1a28c2b80ece0cbfb73b8, =
TABLE
>>                                       =3D> {{NAME =3D> =
'system_mappings',=20
>> FAMILIES =3D> [{NAME =3D> 'f1', BLOOMFILTER =3D> 'NONE', =
REPLICATION_SCO
>>                                       PE =3D> '0', COMPRESSION =3D>=20=

>> 'NONE', VERSIONS =3D> '1', TTL =3D> '2147483647', BLOCKSIZE =3D> =
'65536', IN_ME
>>                                       MORY =3D> 'false', BLOCKCACHE =
=3D>=20
>> 'true'}]}}
>> info:server                             timestamp=3D1318606241185,
>> value=3Dslave2:60020
>> info:serverstartcode       timestamp=3D1318606241185, =
value=3D1316689293206
>> 3 row(s) in 0.3520 seconds
>>=20
>> I tried 'deleteall' command :
>> hbase(main):002:0> deleteall
>> =
'.META.','system_mappings,,1318585995582.bded23240ed1a28c2b80ece0cbfb73b8.=
','info'
>> 0 row(s) in 0.0120 seconds
>>=20
>> Again 'get' command  output :
>> hbase(main):003:0> get
>> =
'.META.','system_mappings,,1318585995582.bded23240ed1a28c2b80ece0cbfb73b8.=
','info'
>> COLUMN                             CELL
>> info:regioninfo                   timestamp=3D1318606241130, =
value=3DREGION =3D>
>> {NAME =3D> 'system_mappings,,1318585995582.bded23240ed1a28c2
>>                                         b80ece0cbfb73b8.', STARTKEY=20=

>> =3D> '', ENDKEY =3D> '', ENCODED =3D> =
bded23240ed1a28c2b80ece0cbfb73b8, TABLE
>>                                        =3D> {{NAME =3D>=20
>> 'system_mappings', FAMILIES =3D> [{NAME =3D> 'f1', BLOOMFILTER =3D> =
'NONE', REPLICATION_SCO
>>                                             PE =3D> '0', COMPRESSION =
=3D>=20
>> 'NONE', VERSIONS =3D> '1', TTL =3D> '2147483647', BLOCKSIZE =3D> =
'65536', IN_ME
>>                                            MORY =3D> 'false',=20
>> BLOCKCACHE =3D> 'true'}]}}
>> info:server                       timestamp=3D1318606241185,
>> value=3Dslave2:60020
>> info:serverstartcode              timestamp=3D1318606241185,
>> value=3D1316689293206
>> 3 row(s) in 0.0130 seconds
>>=20
>>=20
>> -----Original Message-----
>> From: Dejan Menges [mailto:dejan.menges@gmail.com]
>> Sent: Friday, October 14, 2011 6:00 PM
>> To: user@hbase.apache.org
>> Subject: Re: Not able to remove entries from .META. table
>>=20
>> Or you can format your namenode and prior to that manually clean your=20=

>> folders hosting HDFS data.
>>=20
>> On Fri, Oct 14, 2011 at 2:27 PM, Harsh J <harsh@cloudera.com> wrote:
>>=20
>>> Stuti,
>>>=20
>>> What was your deleteall command for deleting the .META. entry =
exactly?
>>> You generally have to pass the whole region ID string to it to=20
>>> remove the entry properly. If you can provide the entry from your=20
>>> .META., I can churn out the exact command you're required to run to=20=

>>> properly remove
>> away the entry.
>>>=20
>>> On 14-Oct-2011, at 5:35 PM, Stuti Awasthi wrote:
>>>=20
>>>> Hi,
>>>> I am stuck in a problem that I created some table which was not=20
>>>> created
>>> properly. Its entries are added in .Meta. table but the regions are=20=

>>> offline so I wanted to delete the table.
>>>> I tried to first delete table with 'disable' and 'drop ' commands=20=

>>>> but the
>>> table is not deleted.
>>>> Then I tried to remove the rows entry containing corrupt table=20
>>>> name from
>>> .META. table using 'delete', 'deleteall' utilities but the row=20
>>> entries are not getting deleted.
>>>>=20
>>>> I also found out that when I 'list' the tables , it shows me the=20
>>>> corrupt
>>> table but in HDFS , there is no directory created with corrupt table
>> name.
>>>>=20
>>>> I tried this again by restarting the Hbase server but same issue.=20=

>>>> I am
>>> able to access all other tables properly.
>>>>=20
>>>> Please help
>>>>=20
>>>>=20
>>>> ::DISCLAIMER::
>>>>=20
>>> --------------------------------------------------------------------
>>> --
>>> -------------------------------------------------
>>>>=20
>>>> The contents of this e-mail and any attachment(s) are confidential=20=

>>>> and
>>> intended for the named recipient(s) only.
>>>> It shall not attach any liability on the originator or HCL or its
>>> affiliates. Any views or opinions presented in
>>>> this email are solely those of the author and may not necessarily=20=

>>>> reflect
>>> the opinions of HCL or its affiliates.
>>>> Any form of reproduction, dissemination, copying, disclosure,
>>> modification, distribution and / or publication of
>>>> this message without the prior written consent of the author of=20
>>>> this
>>> e-mail is strictly prohibited. If you have
>>>> received this email in error please delete it and notify the=20
>>>> sender
>>> immediately. Before opening any mail and
>>>> attachments please check them for viruses and defect.
>>>>=20
>>>>=20
>>> --------------------------------------------------------------------
>>> --
>>> -------------------------------------------------
>>>=20
>>>=20
>>=20
>=20
>=20
>=20
> --
> // Jonathan Hsieh (shay)
> // Software Engineer, Cloudera
> // jon@cloudera.com