Yup, as far as I understand. It is application specific how sophisticated and extensive we want this to be. Unless I am off-base here (Stack or other experts can correct me?)
Regards, Shahab On Wed, Jul 31, 2013 at 10:02 AM, Pavan Sudheendra <[email protected]>wrote: > So a secondary table here is more of a temporary table which I can just do > a get operation and get the value? > > Am I right? > > Regards, > Pavan > On Jul 31, 2013 7:24 PM, "Shahab Yunus" <[email protected]> wrote: > > > Please correct me if I am wrong but I think there is as such no hard and > > fast technique for it. There are no constructs or method to this > > specifically in HBase. Your client while writing have to make sure to > write > > to both tables: 1) the main table 2) and the secondary index table. > > Basically it is up to you according to the use-case and application. You > > can delegate this to a sort of a middle (separate layer from client app > at > > least) which takes care of maintaing consistency between the two tables. > > This layer will be used for both read and write to implement and handle > > your indexing. The one obvious issue that you need to take into > > consideration is atomicity and consistency of the two tables. > > > > Regards, > > Shahab > > > > > > On Wed, Jul 31, 2013 at 9:37 AM, Pavan Sudheendra <[email protected] > > >wrote: > > > > > Yes perfect. Could i get few examples on how to create a secondary > lookup > > > table? > > > > > > > > > On Wed, Jul 31, 2013 at 7:01 PM, Jean-Marc Spaggiari > > > <[email protected]> wrote: > > > > Hi Pavan, > > > > > > > > for your lookup table, you need to lookup on the value? Not on the > > they? > > > If > > > > that's the case, just "revert" your table into a secondary index > lookup > > > > table where the value will be the key, and the key will be the value. > > > That > > > > way you just to a get on this secondary index lookup table instead of > > > doing > > > > a scan... > > > > > > > > Make sense? > > > > > > > > JM > > > > > > > > 2013/7/31 Pavan Sudheendra <[email protected]> > > > > > > > >> Before inserting a row in my main table, i need to do a lookup on my > > > >> lookup table. I'm really not sure what logic has to be behind this. > > > >> Should i do a full table scan of the lookup table every single time > i > > > >> insert in the main table? > > > >> > > > >> Thanks jean > > > >> > > > >> > > > >> > > > >> On Wed, Jul 31, 2013 at 6:10 PM, Jean-Marc Spaggiari > > > >> <[email protected]> wrote: > > > >> > the best way is to insert on your lookup table each time you > insert > > on > > > >> your > > > >> > main table. That way you don't have to scan your entire main table > > to > > > >> build > > > >> > the lookup table... > > > >> > > > > >> > If you want to filter by CF you can look at > > > >> > > > > >> > > > > > > http://hbase.apache.org/apidocs/org/apache/hadoop/hbase/filter/FamilyFilter.html > > > >> > . > > > >> > > > > >> > Basically, you have all the filters listed there: > > > >> > > > > >> > > > > > > http://hbase.apache.org/apidocs/org/apache/hadoop/hbase/filter/package-summary.html > > > >> > > > > >> > You should be able to find one which match your needs. > > > >> > > > > >> > JM > > > >> > > > > >> > 2013/7/31 Pavan Sudheendra <[email protected]> > > > >> > > > > >> >> Hi Jean, > > > >> >> Is there a way to specify column family, column qualifier and > > column > > > >> >> value to get the row key without having to scan the table every > > > single > > > >> >> time? > > > >> >> Basically, i'm building a lookup table which helps in insertion > of > > > >> >> another HBase table. its quite complicated to explain. > > > >> >> > > > >> >> On Wed, Jul 31, 2013 at 5:53 PM, Jean-Marc Spaggiari > > > >> >> <[email protected]> wrote: > > > >> >> > Hi Pavan, > > > >> >> > > > > >> >> > Are you looking for something like > > > >> >> > > > > >> >> > > > >> > > > > > > http://hbase.apache.org/apidocs/org/apache/hadoop/hbase/filter/ValueFilter.html > > > >> >> ? > > > >> >> > > > > >> >> > You are going to do a table full scan... If you value is uniq, > > you > > > >> might > > > >> >> > want to build a secondary index table for direct access... > > > >> >> > > > > >> >> > JM > > > >> >> > > > > >> >> > 2013/7/31 Pavan Sudheendra <[email protected]> > > > >> >> > > > > >> >> >> How do i get the row key by specifying a row value assuming > row > > > >> values > > > >> >> >> are unique? > > > >> >> >> > > > >> >> >> What API should i use for this purpose? > > > >> >> >> -- > > > >> >> >> Regards- > > > >> >> >> Pavan > > > >> >> >> > > > >> >> > > > >> >> > > > >> >> > > > >> >> -- > > > >> >> Regards- > > > >> >> Pavan > > > >> >> > > > >> > > > >> > > > >> > > > >> -- > > > >> Regards- > > > >> Pavan > > > >> > > > > > > > > > > > > -- > > > Regards- > > > Pavan > > > > > >
