[
https://issues.apache.org/jira/browse/SOLR-14022?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16992081#comment-16992081
]
Joel Bernstein edited comment on SOLR-14022 at 12/10/19 3:04 AM:
-----------------------------------------------------------------
There are number of reasons that for the health of the project I think CDCR
should be removed:
1) We are going to have to a get a handle on Solr Clouds stability issues. This
is hard enough to get right within a single Solr Cloud and we certainly have
not achieved this yet. Realistically trying to maintain a resilient CDCR that
lives inside of Solr Cloud, as it does now, may not be possible. Removing CDCR
from the guts of Solr Cloud allows the committers to work on Solr Cloud without
having to also worry about CDCR.
2) CDCR's design was likely flawed from the beginning. I was involved in pretty
much all of the design discussions and I really don't think we got it right.
The design is simply too brittle. It relies on everything working perfectly
under heavy indexing load and shifting cluster state.
3) There are probably more reliable approaches we could replace it with.
Approaches that aren't deeply embedded within Solr Cloud and can operate
properly when Solr Cloud is in flux on either side.
4) I believe in its current state CDCR is more of trap for the community with a
history of operational problems.
5) I don't think there is much desire among committers to really dig in and
support the current implementation of CDCR.
6) In general the project is quite bloated and complex and it is important to
remove features that are problematic. This particular feature is quite large
and very complex. So if there is agreement that's also problematic, then I
think its time to remove it.
was (Author: joel.bernstein):
There are number of reasons that for the health of the project I think CDCR
should be removed:
1) We are going to have to a get a handle on Solr Clouds stability issues. This
is hard enough to get right within a single Solr Cloud and we certainly have
not achieved this yet. Realistically trying to maintain a resilient CDCR that
lives inside of Solr Cloud, as it does now, may not be possible. Removing CDCR
from the guts of Solr Cloud allows the committers to work on Solr Cloud without
having to also worry about CDCR.
2) CDCR's design was likely flawed from the beginning. I was involved in pretty
much all of the design discussions and I really don't think we got it right.
The design is simply too brittle. It relies on everything working perfectly
under heavy indexing load and shifting cluster state.
3) There are probably more reliable approaches we could replace it with.
Approaches that aren't deeply embedded within Solr Cloud and can operate
properly when Solr Cloud is in flux on either side.
4) I believe in its current state CDCR is more of trap for the community with a
history of operational problems.
5) I don't think there is much desire among committers to really dig in and
support the current implementation of CDCR.
6) In general the project is quite bloated and complex and it is important to
remove features that are problematic. This particular feature is quite large
and very complex. So it becomes a candidate for removal based simply on
removing bloat and complexity, especially attached to Solr Cloud.
> Remove CDCR from Solr
> ---------------------
>
> Key: SOLR-14022
> URL: https://issues.apache.org/jira/browse/SOLR-14022
> Project: Solr
> Issue Type: Improvement
> Security Level: Public(Default Security Level. Issues are Public)
> Components: CDCR
> Reporter: Joel Bernstein
> Priority: Major
>
> This ticket will remove CDCR from Solr
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
