[
https://issues.apache.org/jira/browse/HBASE-13408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14708312#comment-14708312
]
Edward Bortnikov commented on HBASE-13408:
------------------------------------------
[~mbertozzi] we absolutely agree that there is a room for more optimizations.
We strived to demonstrate a minimum viable implementation, and this is already
a big patch. The published performance results show benefits for both random
access and scan patterns, for a variety of data distributions. Admittedly, we
were more after speed than memory optimization.
Re/ the concrete questions:
1. HFile implementation for memstore segments. Very plausible (in fact, the
internal API's allow to plug in this implementation quite easily).
2. Automatic memstore compaction (not only in-memory families). Our thinking
was to let the system administrator remain in control of what is happening, so
we introduced this limitation. Could be relaxed in the future.
Thanks for all the feedback. We are open to discussion and adjustments that
would eventually lead this feature to HBase mainstream.
> HBase In-Memory Memstore Compaction
> -----------------------------------
>
> Key: HBASE-13408
> URL: https://issues.apache.org/jira/browse/HBASE-13408
> Project: HBase
> Issue Type: New Feature
> Reporter: Eshcar Hillel
> Attachments: HBASE-13408-trunk-v01.patch,
> HBASE-13408-trunk-v02.patch,
> HBaseIn-MemoryMemstoreCompactionDesignDocument-ver02.pdf,
> HBaseIn-MemoryMemstoreCompactionDesignDocument.pdf,
> InMemoryMemstoreCompactionEvaluationResults.pdf,
> InMemoryMemstoreCompactionScansEvaluationResults.pdf
>
>
> A store unit holds a column family in a region, where the memstore is its
> in-memory component. The memstore absorbs all updates to the store; from time
> to time these updates are flushed to a file on disk, where they are
> compacted. Unlike disk components, the memstore is not compacted until it is
> written to the filesystem and optionally to block-cache. This may result in
> underutilization of the memory due to duplicate entries per row, for example,
> when hot data is continuously updated.
> Generally, the faster the data is accumulated in memory, more flushes are
> triggered, the data sinks to disk more frequently, slowing down retrieval of
> data, even if very recent.
> In high-churn workloads, compacting the memstore can help maintain the data
> in memory, and thereby speed up data retrieval.
> We suggest a new compacted memstore with the following principles:
> 1. The data is kept in memory for as long as possible
> 2. Memstore data is either compacted or in process of being compacted
> 3. Allow a panic mode, which may interrupt an in-progress compaction and
> force a flush of part of the memstore.
> We suggest applying this optimization only to in-memory column families.
> A design document is attached.
> This feature was previously discussed in HBASE-5311.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
