hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Anastasia Braginsky (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-18010) Connect CellChunkMap to be used for flattening in CompactingMemStore
Date Thu, 18 May 2017 13:11:05 GMT

    [ https://issues.apache.org/jira/browse/HBASE-18010?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16015724#comment-16015724
] 

Anastasia Braginsky commented on HBASE-18010:
---------------------------------------------

Thank you [~anoop.hbase]! I get what you are saying, but I still need more guidance to get
to the KeyValue.heapSize(). I am looking on estimatedSerializedSizeOf:

{code}
public static int estimatedSerializedSizeOf(final Cell cell) {
    if (cell instanceof ExtendedCell) {
      return ((ExtendedCell) cell).getSerializedSize(true) + Bytes.SIZEOF_INT;
    }

    return getSumOfCellElementLengths(cell) +
      // Use the KeyValue's infrastructure size presuming that another implementation would
have
      // same basic cost.
      KeyValue.ROW_LENGTH_SIZE + KeyValue.FAMILY_LENGTH_SIZE +
      // Serialization is probably preceded by a length (it is in the KeyValueCodec at least).
      Bytes.SIZEOF_INT;
  }
{code}

I assume we are talking about ExtendedCell so looking on getSerializedSize (of KeyValue).
By the way why integer is added there, near getSerializedSize?

{code}
public int getSerializedSize(boolean withTags) {
    if (withTags) {
      return this.length;
    }
    return this.getKeyLength() + this.getValueLength() + KEYVALUE_INFRASTRUCTURE_SIZE;
  }
{code}

So I do not see KeyValue.heapSize() yet...

> Connect CellChunkMap to be used for flattening in CompactingMemStore
> --------------------------------------------------------------------
>
>                 Key: HBASE-18010
>                 URL: https://issues.apache.org/jira/browse/HBASE-18010
>             Project: HBase
>          Issue Type: Sub-task
>            Reporter: Anastasia Braginsky
>
> The CellChunkMap helps to create a new type of ImmutableSegment, where the index (CellSet's
delegatee) is going to be CellChunkMap. No big cells or upserted cells are going to be supported
here.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message