[
https://issues.apache.org/jira/browse/HDDS-1856?focusedWorklogId=285297&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-285297
]
ASF GitHub Bot logged work on HDDS-1856:
----------------------------------------
Author: ASF GitHub Bot
Created on: 30/Jul/19 22:11
Start Date: 30/Jul/19 22:11
Worklog Time Spent: 10m
Work Description: arp7 commented on issue #1174: HDDS-1856. Make required
changes for Non-HA to use new HA code in OM.
URL: https://github.com/apache/hadoop/pull/1174#issuecomment-516615672
Basically +1 from me. A few minor comments.
I thought about the potential synchronization issue @anuengineer pointed out
offline and am not convinced it exists. Because we hold/release the lock when
flushTransactions calls setReadyBuffer. So that is the synchronization point.
All later access from the same thread should see the correct value of the
futureQueue.
Couple of ways remove the ambiguity:
1. Make the queues volatile.
1. Return the queue from setReadyBuffer. Since the returned pointer was
sampled with the lock held, the caller is guaranteed to see the correct value.
This solution should also make findbugs happy.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 285297)
Time Spent: 2h 50m (was: 2h 40m)
> Make changes required for Non-HA to use new HA code in OM.
> ----------------------------------------------------------
>
> Key: HDDS-1856
> URL: https://issues.apache.org/jira/browse/HDDS-1856
> Project: Hadoop Distributed Data Store
> Issue Type: Sub-task
> Reporter: Bharat Viswanadham
> Assignee: Bharat Viswanadham
> Priority: Major
> Labels: pull-request-available
> Time Spent: 2h 50m
> Remaining Estimate: 0h
>
> In this Jira following things will be implemented:
> # Make the necessary changes for non-HA code path to use Cache and
> DoubleBuffer.
> ## When adding to double buffer, return future. This future will be used in
> the non-HA path to wait for this, and when it is completed return response to
> the client.
> ## Add to double-buffer will happen inside validateAndUpdateCache. In this
> way, in non-HA, when multiple RPC handler threads are calling preExecute and
> validateAndUpdateCache, the order inserted in to double buffer will happen in
> the order requests are received.
>
> In this Jira, we shall not convert non-ha code path to use this, as security
> and acl work is not completed to use this new model.
>
>
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
