[
https://issues.apache.org/jira/browse/HBASE-16583?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15479671#comment-15479671
]
Phil Yang commented on HBASE-16583:
-----------------------------------
Thread pools and queues indeed enlarge the latency of the queries, especially
when the cluster is in normal status. We have to consider the advantage and
disadvantage of introducing them. We indeed meet some issues on slow DNs. For a
online service, .99 or .999 latency depends on GC, and throughput may not be
the most important thing because we can add more nodes. But availability is
absolutely the most important thing for HBase users (and of course, the most
most important is no data-loss..). It is different from offline service. So we
have to avoid any scene that one dead/slow server leads to the availability
reduce more than 1/n for this cluster.
> Staged Event-Driven Architecture
> --------------------------------
>
> Key: HBASE-16583
> URL: https://issues.apache.org/jira/browse/HBASE-16583
> Project: HBase
> Issue Type: Umbrella
> Reporter: Phil Yang
>
> Staged Event-Driven Architecture (SEDA) splits request-handling logic into
> several stages, each stage is executed in a thread pool and they are
> connected by queues.
> Currently, in region server we use a thread pool to handle requests from
> client. The number of handlers is configurable, reading and writing use
> different pools. The current architecture has two limitations:
> Performance:
> Different part of the handling path has different bottleneck. For example,
> accessing MemStore and cache mainly consumes CPU but accessing HDFS mainly
> consumes network/disk IO. If we use SEDA and split them into two different
> stages, we can use different numbers for two pools according to the
> CPU/disk/network performance case by case.
> Availability:
> HBASE-16388 described a scene that if the client use a thread pool and use
> blocking methods to access region servers, only one slow server may exhaust
> most of threads of the client. For HBase, we are the client and HDFS
> datanodes are the servers. A slow datanode may exhaust most of handlers. The
> best way to resolve this issue is make HDFS requests non-blocking, which is
> exactly what SEDA does.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
