[
https://issues.apache.org/jira/browse/TS-4026?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15668503#comment-15668503
]
Bryan Call commented on TS-4026:
--------------------------------
[~bknight]
We have opted to remove clustering in favor of using parent caching or the carp
plugin. Committers on the project are running very large caches where there is
only one object stored across servers using these two techniques. We felt it
was easier to maintain using HTTP and consistent hashing to achieve what cache
clustering was doing using a similar code path as the normal use case, standard
proxying.
We asked for input on features we were thinking of removing before the ATS
7.0.0 release and reversed the decision on two of these because people in the
ATS Community voiced their opinion against removing these features. No one
spoke up for keeping cache clustering and we believed no one was using it. We
would have been very welcome to hear your input at that time.
If you need help on moving to parent caching or using the carp plugin please
email the users mailing list and we will help you.
> Remove clustering feature in 7.0.0 release
> -------------------------------------------
>
> Key: TS-4026
> URL: https://issues.apache.org/jira/browse/TS-4026
> Project: Traffic Server
> Issue Type: Improvement
> Components: Clustering
> Reporter: Bryan Call
> Assignee: Phil Sorber
> Labels: incompatible
> Fix For: 7.0.0
>
> Time Spent: 3h 20m
> Remaining Estimate: 0h
>
> From the ATS Fall Summit 2015 no one said they where using the clustering
> feature.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
