ACCUMULO-4752 Create documentation on improving performance #46
Conversation
mikewalch
changed the title
mikewalch
force-pushed
the
accumulo-4752
branch
from
9a12c4d
to
bf46466
Compare
|
|
||
| 1. Decrease the [major compaction ratio][compaction] of a table to decrease the number of | ||
| files per tablet. Less files reduces the latency of reads. | ||
|
|
There was a problem hiding this comment.
Adjusting table.file.compress.blocksize and table.file.compress.blocksize.index can also be helpful. Lowering table.file.compress.blocksize can result in better random seek performance. However it increases the index size in the file. If the indexes are too large to fit in cache, this can hinder performance. Also, as the index size increases the depth of the index tree in each file may increase. Increasing table.file.compress.blocksize.index can reduce the depth of the tree.
|
|
||
| 1. Increase the [major compaction ratio][compaction] of a table to limit the number of major compactions | ||
| which improves ingest peformance. | ||
|
|
There was a problem hiding this comment.
Setting table.file.compress.type=snappy can increase write spead at the expense of using more disk space.
* Created documentation on RFile along with diagram
|
@mikewalch the picture looks great. A metablock contains all of the the root nodes for each locality group. So this is slightly off in the picture. |
|
@keith-turner, I updated the picture. Let me know if that works. |
ACCUMULO-4752 Create documentation on improving performance (#46) * Also, created documentation on RFile along with diagram
No description provided.