[
https://issues.apache.org/jira/browse/HDFS-7858?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14643102#comment-14643102
]
Arpit Agarwal commented on HDFS-7858:
-------------------------------------
I am +1 on the v11 patch. {{RequestHedgingProxyProvider}} is disabled by
default so remaining issues can be addressed separately to avoid spinning on
this forever. :-)
# One optimization with your new approach - In the common HA case with two
NameNodes, after {{performFailover}} is called, {{toIgnore}} will be non-null.
We don't need to create a thread pool/completion service, we can simply send
the request to the single proxy in the callers thread.
# The TODO is not technically a TODO. We can just document this property in the
class Javadoc that it can block indefinitely and depends on the caller
implementing a timeout.
# Couple of documentation nitpicks:
## _The two implementations which currently ships_ -> _The two implementations
which currently ship_
## _so use these_ --> _so use one of these unless you are using a custom proxy
provider_
Will hold off committing in case [~jingzhao] has any further comments. Thanks
for working on this [~asuresh].
> Improve HA Namenode Failover detection on the client
> ----------------------------------------------------
>
> Key: HDFS-7858
> URL: https://issues.apache.org/jira/browse/HDFS-7858
> Project: Hadoop HDFS
> Issue Type: Improvement
> Reporter: Arun Suresh
> Assignee: Arun Suresh
> Labels: BB2015-05-TBR
> Attachments: HDFS-7858.1.patch, HDFS-7858.10.patch,
> HDFS-7858.10.patch, HDFS-7858.11.patch, HDFS-7858.2.patch, HDFS-7858.2.patch,
> HDFS-7858.3.patch, HDFS-7858.4.patch, HDFS-7858.5.patch, HDFS-7858.6.patch,
> HDFS-7858.7.patch, HDFS-7858.8.patch, HDFS-7858.9.patch
>
>
> In an HA deployment, Clients are configured with the hostnames of both the
> Active and Standby Namenodes.Clients will first try one of the NNs
> (non-deterministically) and if its a standby NN, then it will respond to the
> client to retry the request on the other Namenode.
> If the client happens to talks to the Standby first, and the standby is
> undergoing some GC / is busy, then those clients might not get a response
> soon enough to try the other NN.
> Proposed Approach to solve this :
> 1) Since Zookeeper is already used as the failover controller, the clients
> could talk to ZK and find out which is the active namenode before contacting
> it.
> 2) Long-lived DFSClients would have a ZK watch configured which fires when
> there is a failover so they do not have to query ZK everytime to find out the
> active NN
> 2) Clients can also cache the last active NN in the user's home directory
> (~/.lastNN) so that short-lived clients can try that Namenode first before
> querying ZK
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
