[
https://issues.apache.org/jira/browse/IGNITE-21867?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Alexander Lapin updated IGNITE-21867:
-------------------------------------
Description:
h3. Motivation
RPC_TIMEOUT was mistakenly recognized as a network layer timeout, but actually
it's a business operation one. Within the context of replicaService, there are
legitimate reasons for operation to be processed longer than current timeout of
3 seconds, basically it should be possible to wait indefinitely until the
transaction times out. But because we don't have transaction timeout and
because of possible bugs it seems reasonable to have an operation processing
timeout "watchdog" with finite but relatively big configurable value.
h3. Definition of Done
* Add ability to configure ReplicaService#RPC_TIMEOUT and
TxMessageSender#RPC_TIMEOUT and increase the default values
was:
h3. Motivation
RPC_TIMEOUT was mistakenly recognized as a network layer timeout, but actually
it's a business operation one. Within the context of replicaService, there are
legitimate reasons for operation to be processed longer than current timeout of
3 seconds, basically it should be possible to wait indefinitely until the
transaction times out. But because we don't have transaction timeout and
because of possible bugs it seems reasonable to have an operation processing
timeout "watchdog" with finite but relatively big configurable value.
h3. Definition of Done
> Add ability to configure ReplicaService#RPC_TIMEOUT and
> TxMessageSender#RPC_TIMEOUT and increase the default values
> -------------------------------------------------------------------------------------------------------------------
>
> Key: IGNITE-21867
> URL: https://issues.apache.org/jira/browse/IGNITE-21867
> Project: Ignite
> Issue Type: Improvement
> Reporter: Alexander Lapin
> Priority: Major
>
> h3. Motivation
> RPC_TIMEOUT was mistakenly recognized as a network layer timeout, but
> actually it's a business operation one. Within the context of replicaService,
> there are legitimate reasons for operation to be processed longer than
> current timeout of 3 seconds, basically it should be possible to wait
> indefinitely until the transaction times out. But because we don't have
> transaction timeout and because of possible bugs it seems reasonable to have
> an operation processing timeout "watchdog" with finite but relatively big
> configurable value.
> h3. Definition of Done
> * Add ability to configure ReplicaService#RPC_TIMEOUT and
> TxMessageSender#RPC_TIMEOUT and increase the default values
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
