[
https://issues.apache.org/jira/browse/CASSANDRA-17431?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17505085#comment-17505085
]
Ekaterina Dimitrova edited comment on CASSANDRA-17431 at 3/11/22, 7:12 PM:
---------------------------------------------------------------------------
- phi_convict_threshold is actually time but I don't think we should touch it
considering its nature.
- memtable_cleanup_threshold is marked deprecated in cassandra.yaml, so should
be just marked also internally in Config as Deprecated I guess
- native_transport_max_concurrent_requests_in_bytes_per_ip and
native_transport_max_concurrent_requests_in_bytes - both have assigned same
value if they are set <=0 so easy transfer to the new framework IMO
- These ones should be transferred too -
native_transport_receive_queue_capacity_in_bytes,
user_defined_function_warn_timeout, user_defined_function_fail_timeout,
validation_preview_purge_head_start_in_sec
- Unfortunately block_for_peers_timeout_in_secs is a problematic one. I will
have to think about it more because of the -1 used there. We can't just migrate
it to "0 means disabled" as 0 already has another meaning. More in the java doc
in Config class
- _repair_request_timeout_in_ms_ - is a new one. Checking with the author.
CC [~maedhroz] , [~dcapwell] and [~mck]
Just FYI
was (Author: e.dimitrova):
- phi_convict_threshold is actually time but I don't think we should touch it
considering its nature.
- memtable_cleanup_threshold is marked deprecated in cassandra.yaml, so should
be just marked also internally in Config as Deprecated I guess
- native_transport_max_concurrent_requests_in_bytes_per_ip and
native_transport_max_concurrent_requests_in_bytes - both have assigned same
value if they are set <=0 so easy transfer to the new framework IMO
- These ones should be transferred too -
native_transport_receive_queue_capacity_in_bytes,
user_defined_function_warn_timeout, user_defined_function_fail_timeout,
validation_preview_purge_head_start_in_sec
- Unfortunately block_for_peers_timeout_in_secs is a problematic one. I will
have to think about because of the -1 used there. We can't just migrate it to
"0 means disabled" as 0 already has another meaning. More in the java doc in
Config class
- _repair_request_timeout_in_ms_ - is a new one. Checking with the author.
CC [~maedhroz] , [~dcapwell] and [~mck]
Just FYI
> Move the rest of the Config parameters to the new Config framework
> ------------------------------------------------------------------
>
> Key: CASSANDRA-17431
> URL: https://issues.apache.org/jira/browse/CASSANDRA-17431
> Project: Cassandra
> Issue Type: Task
> Components: Local/Config
> Reporter: Ekaterina Dimitrova
> Assignee: Ekaterina Dimitrova
> Priority: Normal
> Fix For: 4.x
>
>
> Migrate also all Config parameters that are in Config, but not presented in
> cassandra.yaml to the new config framework. Not presented in the yaml as they
> are considered advanced only for advanced users.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
