HBase
  1. HBase
  2. HBASE-847 new API: HTable.getRow with numVersion specified
  3. HBASE-31

Add means of getting the timestamps for all cell versions: e.g. long [] getVersions(row, column)

    Details

    • Type: Sub-task Sub-task
    • Status: Resolved
    • Priority: Trivial Trivial
    • Resolution: Not a Problem
    • Affects Version/s: 0.2.0
    • Fix Version/s: None
    • Component/s: Client
    • Labels:
      None

      Description

      Should be means of asking hbase for list of all the timestamps associated with a particular cell. The brute force way would be adding a getVersions method but perhaps we can come up w/ something more elegant than this?

        Issue Links

          Activity

          Hide
          stack added a comment -

          Perhaps a getRowTimestamps would be better? It would return a list of all columns in a given row and all timestamps for a given cell?

          Map<Text, long []> getRowTimestamps(Text row);
          
          Show
          stack added a comment - Perhaps a getRowTimestamps would be better? It would return a list of all columns in a given row and all timestamps for a given cell? Map<Text, long []> getRowTimestamps(Text row);
          Hide
          Edward J. Yoon added a comment -

          +1 to "getVersions(row, column) "

          Show
          Edward J. Yoon added a comment - +1 to "getVersions(row, column) "
          Hide
          Bryan Duxbury added a comment -

          I was hoping for a method that would give you all of the timestamps for a row, not per row and column. You'd want to able to do this so that you could know what different versions of a particular entity you have in HBase. As such, basically all I'm looking for is:

          <code>long[] getRowTimestamps(Text row)</code>

          Show
          Bryan Duxbury added a comment - I was hoping for a method that would give you all of the timestamps for a row, not per row and column. You'd want to able to do this so that you could know what different versions of a particular entity you have in HBase. As such, basically all I'm looking for is: <code>long[] getRowTimestamps(Text row)</code>
          Hide
          Bryan Duxbury added a comment -

          It seems like after HADOOP-1550 all that really needs to be done to accomplish getRowTimestamps is to consume HRegion.getKeys cleverly.

          I think we should support:

          /** 
           * returns descending list of all row timestamps
           */
          long[] getRowTimestamps(Text row);
          

          @Edward - what would getVersions do? When would you want to know what the versions of a particular column and not the entire row?

          Show
          Bryan Duxbury added a comment - It seems like after HADOOP-1550 all that really needs to be done to accomplish getRowTimestamps is to consume HRegion.getKeys cleverly. I think we should support: /** * returns descending list of all row timestamps */ long [] getRowTimestamps(Text row); @Edward - what would getVersions do? When would you want to know what the versions of a particular column and not the entire row?
          Hide
          Edward J. Yoon added a comment -

          @Bryan - I think It will be useful for time series analysis.

          Show
          Edward J. Yoon added a comment - @Bryan - I think It will be useful for time series analysis.
          Hide
          Edward J. Yoon added a comment - - edited

          For example,

          • Financial engineering
          • Informative dimension and Time dimension factors in Search engine Rank algorithm. (but, i like the entropy theory)
          • ...etc.
          Show
          Edward J. Yoon added a comment - - edited For example, Financial engineering Informative dimension and Time dimension factors in Search engine Rank algorithm. (but, i like the entropy theory) ...etc.
          Hide
          Edward J. Yoon added a comment -

          If stack is ok, i'll do it.

          Show
          Edward J. Yoon added a comment - If stack is ok, i'll do it.
          Hide
          Edward J. Yoon added a comment -

          >> When would you want to know what the versions of a particular column and not the entire row?

          Yeap. Need a time series of a particular (row, column).

          Show
          Edward J. Yoon added a comment - >> When would you want to know what the versions of a particular column and not the entire row? Yeap. Need a time series of a particular (row, column).
          Hide
          Bryan Duxbury added a comment -

          What is the correct type of Writable to use to transmit a List<long>?

          Show
          Bryan Duxbury added a comment - What is the correct type of Writable to use to transmit a List<long>?
          Hide
          stack added a comment -

          Will an array of longs do?

          Show
          stack added a comment - Will an array of longs do?
          Hide
          Bryan Duxbury added a comment -

          Sure, long[] will work.

          Show
          Bryan Duxbury added a comment - Sure, long[] will work.
          Hide
          Bryan Duxbury added a comment -

          I think that perhaps this issue won't be of as much importance once gets return the relevant timestamp with the values returned.

          Show
          Bryan Duxbury added a comment - I think that perhaps this issue won't be of as much importance once gets return the relevant timestamp with the values returned.
          Hide
          stack added a comment -

          Demoting priority even further. Let some one ask for it before we'll even consider going near this issue. For now, focus will be permutations that return timestamp+data always.

          Show
          stack added a comment - Demoting priority even further. Let some one ask for it before we'll even consider going near this issue. For now, focus will be permutations that return timestamp+data always.
          Hide
          stack added a comment -

          Moving to 0.20.0 with Doğacan Güney's permission over in hbase-847

          Show
          stack added a comment - Moving to 0.20.0 with Doğacan Güney's permission over in hbase-847
          Hide
          Harsh J added a comment -

          This can be resolved, given Bryan's comments earlier.

          I think that perhaps this issue won't be of as much importance once gets return the relevant timestamp with the values returned.

          This is available now.

          Show
          Harsh J added a comment - This can be resolved, given Bryan's comments earlier. I think that perhaps this issue won't be of as much importance once gets return the relevant timestamp with the values returned. This is available now.
          Hide
          stack added a comment -

          Thanks Harsh for doing the janitorial cleanups

          Show
          stack added a comment - Thanks Harsh for doing the janitorial cleanups
          Hide
          stack added a comment -

          I think this issue was asking for return of timestamps only, not timestamps + data. I think its good you closed it though Harsh; its an old issue that hasn't been worked on in three years.

          Show
          stack added a comment - I think this issue was asking for return of timestamps only, not timestamps + data. I think its good you closed it though Harsh; its an old issue that hasn't been worked on in three years.

            People

            • Assignee:
              Doğacan Güney
              Reporter:
              stack
            • Votes:
              1 Vote for this issue
              Watchers:
              1 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved:

                Development