[
https://issues.apache.org/jira/browse/IGNITE-18134?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Alexander Lapin updated IGNITE-18134:
-------------------------------------
Description:
h3. Motivation
[IGNITE-18121|https://issues.apache.org/jira/browse/IGNITE-18121] introduced
DATA_NODES_AUTO_ADJUST_SCALE_UP scheduler and
[IGNITE-18132|https://issues.apache.org/jira/browse/IGNITE-18132] SCALE_DOWN
one. Let's take a look at the last remaining auto adjust property -
DATA_NODES_AUTO_ADJUST. The reason we have one is to eliminate excessive
rebalance in case of users intention on having the same value for both scale_up
and scale_down. As we saw in the example above, the events of adding and
removing nodes fall into the corresponding frames with a dedicated timer each:
one for expanding the topology (adding nodes), and another for narrowing it
(removing nodes), which in turn leads to two rebalances - one per each frame.
If the user however wants to put both types of events (adding and removing
nodes) in one frame with only one dataNodes recalculation and one rebalance, he
should use the DATA_NODES_AUTO_ADJUST property.
h3. Definition of Done
DataNodes recalculation transitively triggered by both node appeared and
disappeared events is delayed for dataNodesAutoAdjust value.
In case of new logical topology event, existing auto adjust timer should be
re-scheduled.
h3. Implementation Notes
Same as for IGNITE-18121 and IGNITE-18132.
> Introduce common auto adjust scheduler
> --------------------------------------
>
> Key: IGNITE-18134
> URL: https://issues.apache.org/jira/browse/IGNITE-18134
> Project: Ignite
> Issue Type: Improvement
> Reporter: Alexander Lapin
> Priority: Major
> Labels: ignite-3
>
> h3. Motivation
> [IGNITE-18121|https://issues.apache.org/jira/browse/IGNITE-18121] introduced
> DATA_NODES_AUTO_ADJUST_SCALE_UP scheduler and
> [IGNITE-18132|https://issues.apache.org/jira/browse/IGNITE-18132] SCALE_DOWN
> one. Let's take a look at the last remaining auto adjust property -
> DATA_NODES_AUTO_ADJUST. The reason we have one is to eliminate excessive
> rebalance in case of users intention on having the same value for both
> scale_up and scale_down. As we saw in the example above, the events of adding
> and removing nodes fall into the corresponding frames with a dedicated timer
> each: one for expanding the topology (adding nodes), and another for
> narrowing it (removing nodes), which in turn leads to two rebalances - one
> per each frame. If the user however wants to put both types of events (adding
> and removing nodes) in one frame with only one dataNodes recalculation and
> one rebalance, he should use the DATA_NODES_AUTO_ADJUST property.
> h3. Definition of Done
> DataNodes recalculation transitively triggered by both node appeared and
> disappeared events is delayed for dataNodesAutoAdjust value.
> In case of new logical topology event, existing auto adjust timer should be
> re-scheduled.
> h3. Implementation Notes
> Same as for IGNITE-18121 and IGNITE-18132.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
