Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...


Secondary index groups are a breaking change, and other index implementations will need modification for compatibility.

New JMX metrics

Users will have access to extensive JMX metrics to help them isolate performance issues on their clusters.

There are enough that listing them here would be counterproductive; however they fall into the following groups:

  • Index - Metrics specific to a single index.
  • Column query - Metrics specific to an indexed column.
    • Trie index - Low level metrics for trie indexes.
    • KD-Tree - Low level metrics for kd-tree indexes.
  • Index group - Metrics for shared resources when multiple columns are indexes on the same table.
  • Table query - Metrics for all queries on a single table.
    • Per query - Metrics for queries.

Metrics include items such as: disk usage, memory usage, query latencies, compaction statistics, chunk cache hits/misses/lookups, open files.

Compatibility, Deprecation, and Migration Plan

...

The following applies to the version 1 index format.. There is a version 2 index format under development.

JMX metrics

Users will have access to extensive JMX metrics to help them isolate performance issues on their clusters.

There are enough that listing them here would be counterproductive; however they fall into the following groups:

  • Index - Metrics specific to a single index.
  • Column query - Metrics specific to an indexed column.
    • Trie index - Low level metrics for trie indexes.
    • KD-Tree - Low level metrics for kd-tree indexes.
  • Index group - Metrics for shared resources when multiple columns are indexes on the same table.
  • Table query - Metrics for all queries on a single table.
    • Per query - Metrics for queries.


Metrics include items such as: disk usage, memory usage, query latencies, compaction statistics, chunk cache hits/misses/lookups, open files.

Terminology

  • Row ID - A monotonic increasing integer associated with every row in a sstable. It’s stored in an index structure instead of key token or key offset, because it compresses better.
  • Postings/posting-list - Sorted row ids that match a given indexed value. 
  • Token file - An index of Row ID -> partition key token for every row in the sstable.
  • Offset file - An index of Row ID -> partition key offset on the data/primary-index file for every row in the sstable.
  • Segment - A smallest unit of on-disk indexing structure that is flushed during compaction to reduce memory pressure. Multiple segments of an index are written to the same physical file.

...