[JBoss JIRA] (ISPN-8740) Refactor backup-write commands (triangle)
by Radim Vansa (JIRA)
[ https://issues.jboss.org/browse/ISPN-8740?page=com.atlassian.jira.plugin.... ]
Radim Vansa edited comment on ISPN-8740 at 1/31/18 9:46 AM:
------------------------------------------------------------
I prefer to decouple backup commands from regular commands, that's why I've suggested to link them through helpers rather than adding extra methods to interfaces.
These helpers also keep all the lambdas together, so once you start passing >= 3 parameters ([like here|https://github.com/infinispan/infinispan/commit/74f4c8962cfad0e406e1...]) it makes passing these more compact. Not that it would be applicable everywhere.
was (Author: rvansa):
I prefer to decouple backup commands from regular commands, that's why I've suggested to link them through helpers rather than adding extra methods to interfaces.
These helpers also keep all the lambdas together, so once you start passing > 3 parameters ([like here|https://github.com/infinispan/infinispan/commit/74f4c8962cfad0e406e1...]) it makes passing these more compact. Not that it would be applicable everywhere.
> Refactor backup-write commands (triangle)
> -----------------------------------------
>
> Key: ISPN-8740
> URL: https://issues.jboss.org/browse/ISPN-8740
> Project: Infinispan
> Issue Type: Enhancement
> Reporter: Pedro Ruivo
> Assignee: Pedro Ruivo
>
> h2. Split the backup commands
> * single key
> * single key, functional
> * put map
> * entries functional
> * keys functional
> h2. Remove method from {{WriteCommand}} interface (Radim's suggestion)
> * Use helper methods to create the backup-write commands. (note. I'm not sure if it makes much sense)
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
6 years, 11 months
[JBoss JIRA] (ISPN-8580) Refactor ThreadLocal use to handle hotrod flags
by Katia Aresti (JIRA)
[ https://issues.jboss.org/browse/ISPN-8580?page=com.atlassian.jira.plugin.... ]
Katia Aresti updated ISPN-8580:
-------------------------------
Summary: Refactor ThreadLocal use to handle hotrod flags (was: Reactor ThreadLocal use to handle hotrod flags )
> Refactor ThreadLocal use to handle hotrod flags
> -------------------------------------------------
>
> Key: ISPN-8580
> URL: https://issues.jboss.org/browse/ISPN-8580
> Project: Infinispan
> Issue Type: Task
> Components: Hot Rod
> Reporter: Katia Aresti
> Assignee: Katia Aresti
> Priority: Optional
>
> OperationsFactory and MultimapOperationsFactory use ThreadLocal to share flags.
> Instead store the flags on each instance of the RemoteMultiMap/RemoteCache and pass them when they invoke something. When a new flag is added or removed we would create a new instance. This seems much more intuitive to me and simpler.
> Then the flags can be defaulted on the RemoteCache/RemoteMultiMap when it is retrieved from the manager as well.
>
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
6 years, 11 months
[JBoss JIRA] (ISPN-8740) Refactor backup-write commands (triangle)
by Pedro Ruivo (JIRA)
Pedro Ruivo created ISPN-8740:
---------------------------------
Summary: Refactor backup-write commands (triangle)
Key: ISPN-8740
URL: https://issues.jboss.org/browse/ISPN-8740
Project: Infinispan
Issue Type: Enhancement
Reporter: Pedro Ruivo
Assignee: Pedro Ruivo
h2. Split the backup commands
* single key
* single key, functional
* put map
* entries functional
* keys functional
h2. Remove method from {{WriteCommand}} interface (Radim's suggestion)
* Use helper methods to create the backup-write commands. (note. I'm not sure if it makes much sense)
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
6 years, 11 months
[JBoss JIRA] (ISPN-7765) Allow HotRodDecoder to be shared between Netty pipelines
by Radim Vansa (JIRA)
[ https://issues.jboss.org/browse/ISPN-7765?page=com.atlassian.jira.plugin.... ]
Radim Vansa closed ISPN-7765.
-----------------------------
Resolution: Rejected
> Allow HotRodDecoder to be shared between Netty pipelines
> --------------------------------------------------------
>
> Key: ISPN-7765
> URL: https://issues.jboss.org/browse/ISPN-7765
> Project: Infinispan
> Issue Type: Enhancement
> Reporter: Sebastian Łaskawiec
> Assignee: William Burns
>
> Currently {{HotRodDecoder}} instance can not be shared between Netty pipelines as {{HotRodEncoder}}. Currently each pipeline creates its own instance by calling:
> {code}
> server.getDecoder()
> {code}
> in {{NettyChannelInitializer#initializeChannel}}. We could probably make it a bit more efficient (or limit allocation rate at least) if we allow sharing single instance between pipelines.
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
6 years, 11 months
[JBoss JIRA] (ISPN-7765) Allow HotRodDecoder to be shared between Netty pipelines
by Radim Vansa (JIRA)
[ https://issues.jboss.org/browse/ISPN-7765?page=com.atlassian.jira.plugin.... ]
Radim Vansa commented on ISPN-7765:
-----------------------------------
Actually we can't bind the decoder to the channel as each operation carries Hot Rod version, so you might use different decoders on the same channel (however unlikely that is).
Anyway, I'll close this.
> Allow HotRodDecoder to be shared between Netty pipelines
> --------------------------------------------------------
>
> Key: ISPN-7765
> URL: https://issues.jboss.org/browse/ISPN-7765
> Project: Infinispan
> Issue Type: Enhancement
> Reporter: Sebastian Łaskawiec
> Assignee: William Burns
>
> Currently {{HotRodDecoder}} instance can not be shared between Netty pipelines as {{HotRodEncoder}}. Currently each pipeline creates its own instance by calling:
> {code}
> server.getDecoder()
> {code}
> in {{NettyChannelInitializer#initializeChannel}}. We could probably make it a bit more efficient (or limit allocation rate at least) if we allow sharing single instance between pipelines.
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
6 years, 11 months