hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From lars hofhansl <lhofha...@yahoo.com>
Subject Re: Book & site update
Date Wed, 12 Oct 2011 04:47:26 GMT
If/when HBASE-4536 gets committed with this specific change, I'll add a change for the documentation
as well.

As of now M=N is till valid. It's not wrong, it just does not make sense.

M=N is saying: Keep at most N versions, expire rows after T secs. But keep at least N versions
The effect is that the same as no TTL.

From: Doug Meil <doug.meil@explorysmedical.com>
To: "user@hbase.apache.org" <user@hbase.apache.org>
Sent: Tuesday, October 11, 2011 2:52 PM
Subject: Re: Book & site update

Thanks Ted.  

LarsH did that specific entry in the book...  Lars, can you update the
book with the most current thoughts on that particular subject?

On 10/11/11 4:08 PM, "Ted Yu" <yuzhihong@gmail.com> wrote:

>Thanks Doug for the update.
>In 6.4.2:
>"keep the last T minutes worth of data, at most N versions, *but keep at
>least M versions around*" (where M is the value for minimum number of row
>versions, M<=N)
>Please refer to discussion toward the end of HBASE-4536 where Lars
>dropping support for the case of M==N > 0.
>On Tue, Oct 11, 2011 at 12:59 PM, Doug Meil
>> Hi folks-
>> Team-list.html
>>  *   Welcome LarsH!  Lars H. is the latest committer, and is now on the
>> team-list page.
>> Book
>>  *   Architecture/Regions/Store:  a lot of KeyValue documentation has
>> added on how exactly HBase stores data internally
>>  *   Schema Design:  Versions has been clarified (min/max now separate
>> sub-sections)
>>  *   Schema Design:  Schema Design Smackdown!  A new section has been
>> in this chapter for commonly asked schema design question.  (Spoiler
>>  "more rows")
>>  *   OpsMgt:  A first-cut capacity planning section for storage calcs
>> been added.
>> Doug Meil
>> Chief Software Architect, Explorys
>> doug.meil@explorys.com
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message