accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Adam Fuchs (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ACCUMULO-4039) try out a proactor design pattern for tserver services
Date Tue, 27 Oct 2015 21:40:27 GMT

    [ https://issues.apache.org/jira/browse/ACCUMULO-4039?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14977219#comment-14977219
] 

Adam Fuchs commented on ACCUMULO-4039:
--------------------------------------

The crux of the matter is definitely the unbounded thrift thread pool. I think the current
hand-off to the readahead thread pool works well, and already uses the future mechanism that
would probably be used in a proactor-like approach. The other issue is that we are forced
to have many different threadpools to avoid potential deadlock the way we do things now, and
I'm speculating we might be able to simplify that to possibly a single threadpool for client
services if we put in the right prioritization.

There's a similar problem on the client side of needing a very large number of threads for
reads in order to reduce latency. That might be improved in a similar fashion.

I'm glad other people seem to be interested in this topic, too!

> try out a proactor design pattern for tserver services
> ------------------------------------------------------
>
>                 Key: ACCUMULO-4039
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-4039
>             Project: Accumulo
>          Issue Type: Improvement
>          Components: tserver
>            Reporter: Adam Fuchs
>            Priority: Minor
>
> For large instances (i.e. lots of clients for a given tserver) we create oodles of threads
on the tserver. This makes for difficulty in predicting performance, memory usage, etc. Moreover,
we have operations that recurse, like a server querying itself, that we currently solve by
having separate thread pools for regular table operations and metadata table operations, and
we "disallow" things like an iterator writing to another table. One alternative option would
be to switch to a Proactor pattern: https://en.wikipedia.org/wiki/Proactor_pattern
> The core of this would be to switch to using a selection set rather than a thread per
active connection, and then wrap everything in sessions that make progress in something like
a state model, with states that account for asynchronous communications and remote work.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message