Sergio Bossa

Currently, compaction creates one large batch per compacted data file, which is sub-optimal as it reduces hints efficiency and can amplify corruption cases as it creates a long chain of...

Add lock file

1

Add lock file to prevent opening the journal from multiple JVMs at the same time.

Currently, compaction goes over all data files and compacts them one-by-one on a single thread: given each data file is immutable and independent from each other, multiple threads could compact...

Improve way to control database defrag

2

Currently, database defrag happens automatically when a configurable percentage of wasted disk space in the data files is reached. Given the defrag operation is very I/O intensive, it would be...

See: https://groups.google.com/forum/?fromgroups=#!topic/nimrod-user/CBbj8gL3gfQ

It should be possible to show and navigate metrics history across logs, in order to allow for correlation of events happening in different logs.

Something like Cubism (http://square.github.com/cubism/) could be used to nicely visualize an aggregation of per-log metrics.

Database backup should be implemented as an HTTP endpoint.

Fixed com.netflix.astyanax.connectionpool.exceptions.PoolTimeoutException

3

Previously, a com.netflix.astyanax.connectionpool.exceptions.PoolTimeoutException was occurring when running a test with a number of concurrent threads higher than 2. This was due to MaxBlockedThreadsPerHost not set into com.netflix.jmeter.connections.a6x.AstyanaxConnection.