Also, I have reduced average load by an order of magnitude,
in an effort to get a handle on this, to no effect -- from
upwards of 100K/queries sec cluster-wide to 10K/queries sec
cluster-wide.
> [...] my workload is modest continuous write operations,
> maybe up to 100/sec, of objects typically < 4K in size
> but can be as large as 20MB. Writes happen to both a
> 'urls' table and a 'content' table.
> 'content' table gets the raw content and uses RECORD
> compression. 'urls' table gets metadata only.
> Concurrent with this are two mapred tasks, one running on
> the 'urls' table, one on the 'content'
> table. The mapred tasks run once every few minutes for a few
> minutes, with a interval between executions currently at 5
> minutes.
> - Andy
