On 24/05/2011 07:51, Norman wrote:
2. If we store each folder in a file, we may have less performance
issue on read (larger file), but we face the issue that we can not
alter the content (only append!!). So does not sound like an option.
Well we could just have some kind of info which mails are deleted and
skip then while read from the file. This would still need to cleanup
"deleted" messages later somehow. Not sure if it makes sense
given by the complexibilty it will introduce..
Yep, I also thought to maintain a list a "expunged/deleted" mails per
mailbox, but that's not the most performant solution.
It's true that the SequenceFile [1] only allows append, the MapWritable
[2] implement java.util.Map, so you've got the put, get, remove...
If we have a MapWritable per Mailbox, we will need to open/close it
frequently (based on user SELECT), this may be not performant (don't
know?). Also, with this approach, we are more in a KeyValue storage
approach, and we may better finally take a real KeyValue store to get
all needed functionality (scan,...).
Tks,
- Eric
[1]
http://hadoop.apache.org/common/docs/current/api/org/apache/hadoop/io/SequenceFile.Writer.html
[2]
http://hadoop.apache.org/common/docs/current/api/org/apache/hadoop/io/MapWritable.html
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
