You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hbase.apache.org by 庄阳 <zh...@asiainfo-linkage.com> on 2011/12/29 02:12:06 UTC
How to solve the problem of bulk load data is overwritten
Hi,
We use bulk load to load multiple records for the same rowkey while in storage, there will be coverage of the situation, even if a large set versions.
There will be new data to overwrite the old data.Do not know how to solve?
2011-12-29
zhuangyang
Re: How to solve the problem of bulk load data is overwritten
Posted by Andrew Purtell <ap...@yahoo.com>.
The question is not clear.
Best regards,
- Andy
On Dec 28, 2011, at 5:12 PM, "庄阳" <zh...@asiainfo-linkage.com> wrote:
> Hi,
> We use bulk load to load multiple records for the same rowkey while in storage, there will be coverage of the situation, even if a large set versions.
> There will be new data to overwrite the old data.Do not know how to solve?
>
> 2011-12-29
>
>
>
> zhuangyang
>
Re: Re: How to solve the problem of bulk load data is overwritten
Posted by Suraj Varma <sv...@gmail.com>.
I'm interpreting your question as "I bulk loaded multiple versions of
a row, but when I issue a get I only get one version back ...".
If so - use Get#setMaxVersions() api to the required number to get
multiple versions back.
If the above interpretation is wrong ... please clarify what you were
expecting to get and what you are getting after the bulk load.
--Suraj
On Wed, Dec 28, 2011 at 7:28 PM, 庄阳 <zh...@asiainfo-linkage.com> wrote:
> Hi,
>
> This is data file:
> 201112;22;145006152;15659424176;460016010360000;147B;2;9;1;-1;20111222100254;20111222100603;591;10487041;0
> 201112;22;144007120;15659437430;460016010302326;9D81;2;9;1;-1;20111222095630;20111222100441;591;10486325;0
> 201112;22;150924671;15659424176;460015902701829;4CB1;2;9;1;-1;20111222090611;20111222100611;591;1284851;0
> 201112;22;145263466;15659437430;460015702708566;3CFD;2;9;1;-1;20111222094056;20111222100611;591;2804698;0
> 201112;22;150307120;15659424176;460015702749026;230D;2;9;1;-1;20111222091923;20111222100613;591;7361;0
> 201112;22;150662025;15659437430;460015912701129;0E4D;2;9;1;-1;20111222094509;20111222100420;591;10486712;0
> 201112;22;145869497;15659424176;460016010367104;0A20;2;9;1;-1;20111222095057;20111222100624;591;2544333;0
> 201112;22;150688057;15659437430;460015702736925;060F;2;9;1;-1;20111222100153;20111222100433;591;10486977;0
>
>
> We used the fourth field as rowkey.Bulk load using the above data into hbase, use the command to view the data is only one version.
>
> We know that hbase three versions of the data stored by default,so create a table we have set versions is 5.
>
>
> 2011-12-29
>
>
>
> zhuangyang
>
>
>
> 发件人: Andrew Purtell
> 发送时间: 2011-12-29 09:57:21
> 收件人: user@hbase.apache.org
> 抄送:
> 主题: Re: How to solve the problem of bulk load data is overwritten
>
> The question is not clear.
> Best regards,
> - Andy
> On Dec 28, 2011, at 5:12 PM, "庄阳" <zh...@asiainfo-linkage.com> wrote:
>> Hi,
>> We use bulk load to load multiple records for the same rowkey while in storage, there will be coverage of the situation, even if a large set versions.
>> There will be new data to overwrite the old data.Do not know how to solve?
>>
>> 2011-12-29
>>
>>
>>
>> zhuangyang
>>
Re: Re: How to solve the problem of bulk load data is overwritten
Posted by 庄阳 <zh...@asiainfo-linkage.com>.
Hi,
This is data file:
201112;22;145006152;15659424176;460016010360000;147B;2;9;1;-1;20111222100254;20111222100603;591;10487041;0
201112;22;144007120;15659437430;460016010302326;9D81;2;9;1;-1;20111222095630;20111222100441;591;10486325;0
201112;22;150924671;15659424176;460015902701829;4CB1;2;9;1;-1;20111222090611;20111222100611;591;1284851;0
201112;22;145263466;15659437430;460015702708566;3CFD;2;9;1;-1;20111222094056;20111222100611;591;2804698;0
201112;22;150307120;15659424176;460015702749026;230D;2;9;1;-1;20111222091923;20111222100613;591;7361;0
201112;22;150662025;15659437430;460015912701129;0E4D;2;9;1;-1;20111222094509;20111222100420;591;10486712;0
201112;22;145869497;15659424176;460016010367104;0A20;2;9;1;-1;20111222095057;20111222100624;591;2544333;0
201112;22;150688057;15659437430;460015702736925;060F;2;9;1;-1;20111222100153;20111222100433;591;10486977;0
We used the fourth field as rowkey.Bulk load using the above data into hbase, use the command to view the data is only one version.
We know that hbase three versions of the data stored by default,so create a table we have set versions is 5.
2011-12-29
zhuangyang
发件人: Andrew Purtell
发送时间: 2011-12-29 09:57:21
收件人: user@hbase.apache.org
抄送:
主题: Re: How to solve the problem of bulk load data is overwritten
The question is not clear.
Best regards,
- Andy
On Dec 28, 2011, at 5:12 PM, "庄阳" <zh...@asiainfo-linkage.com> wrote:
> Hi,
> We use bulk load to load multiple records for the same rowkey while in storage, there will be coverage of the situation, even if a large set versions.
> There will be new data to overwrite the old data.Do not know how to solve?
>
> 2011-12-29
>
>
>
> zhuangyang
>