[
https://issues.apache.org/jira/browse/HDDS-17312?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Siyao Meng updated HDDS-17312:
------------------------------
Epic Link: HDDS-17321
> gRPC client read and EC write timeouts are truncated to whole seconds
> ---------------------------------------------------------------------
>
> Key: HDDS-17312
> URL: https://issues.apache.org/jira/browse/HDDS-17312
> Project: Apache Ozone
> Issue Type: Bug
> Components: Ozone Client
> Reporter: Siyao Meng
> Assignee: Siyao Meng
> Priority: Major
> Labels: pull-request-available
>
> {{XceiverClientGrpc}} reads {{ozone.client.read.timeout}}, and
> {{ECXceiverClientGrpc}} reads {{ozone.client.ec.grpc.write.timeout}}, with
> {{getTimeDuration(..., TimeUnit.SECONDS)}}, and applies the result as the
> gRPC deadline in seconds ({{withDeadlineAfter(timeout, TimeUnit.SECONDS)}}).
> A sub-second value such as {{500ms}} becomes 0, so every request to the
> datanode fails immediately with {{DEADLINE_EXCEEDED}}. A value like
> {{1500ms}} becomes 1s.
> The short-circuit client and server ({{XceiverClientShortCircuit}},
> {{XceiverServerDomainSocket}}) already read {{ozone.client.read.timeout}} in
> milliseconds.
> Proposal: keep the timeout in milliseconds and use it as the deadline in
> milliseconds. A value without unit should still mean seconds for these two
> keys, to stay compatible with existing configurations.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
