Merge Feature/cets into master #4063
Conversation
Use cets for stream resumption backend Print cets tables in mongoosectl mnesia info Enable cets on CI Add test for mongooseimctl mnesia info Test ejabberd_sm_cets in ejabberd_sm_SUITE Rename to cets
Otherwise sm_SUITE fails with already_started error
…_by_nonexistent_name
Merge master into cets
This comment was marked as outdated.
This comment was marked as outdated.
Disable mnesia in pgsql_cets preset
This comment was marked as outdated.
This comment was marked as outdated.
Update CETS dependency
This comment was marked as outdated.
This comment was marked as outdated.
|
small_tests_24 / small_tests / 48b9569 elasticsearch_and_cassandra_25 / elasticsearch_and_cassandra_mnesia / 48b9569 small_tests_25 / small_tests / 48b9569 small_tests_25_arm64 / small_tests / 48b9569 ldap_mnesia_24 / ldap_mnesia / 48b9569 dynamic_domains_pgsql_mnesia_24 / pgsql_mnesia / 48b9569 dynamic_domains_pgsql_mnesia_25 / pgsql_mnesia / 48b9569 ldap_mnesia_25 / ldap_mnesia / 48b9569 dynamic_domains_mysql_redis_25 / mysql_redis / 48b9569 dynamic_domains_mssql_mnesia_25 / odbc_mssql_mnesia / 48b9569 pgsql_cets_25 / pgsql_cets / 48b9569 pgsql_mnesia_24 / pgsql_mnesia / 48b9569 mysql_redis_25 / mysql_redis / 48b9569 internal_mnesia_25 / internal_mnesia / 48b9569 mssql_mnesia_25 / odbc_mssql_mnesia / 48b9569 pgsql_mnesia_25 / pgsql_mnesia / 48b9569 |
There was a problem hiding this comment.
Ok, with such a gigantic diff, I'm not going to carefully read and verify everything changed here, some sections I'll only skim over, trusting that previous rounds of review and tests went well.
Notes:
- I've checked documentation, seems good to me. Not extensive enough but what is there is fine.
- I'm not verifying that everything mnesia now has a cets alternative, I trust having disabled mnesia in tests checked that already.
- I'm not paying thorough attention to pieces like bosh, jingle, or regular muc, I've only read the main module and it seems good, I trust their respective backends and/or helper modules are cool. Though btw, if we already created a dir for
mucwith the_online_helpers, we might have as well putmod_mucin there. Not important, just mentioning. - Changes to ejabberd_router, components, and s2s, they're amazing, thank you!
Some comments request small changes, specially the one in ejabberd_app:start/2, please fix that one 🙏🏽
| @@ -80,6 +80,7 @@ | |||
| {cache_tab, "1.0.30"}, | |||
| {segmented_cache, "0.3.0"}, | |||
| {worker_pool, "6.0.1"}, | |||
| {cets, {git, "https://github.com/esl/cets.git", {branch, "main"}}}, | |||
There was a problem hiding this comment.
I'm going to insist in creating a hex package for this soon.
There was a problem hiding this comment.
Yes, we will do before release (or you could try to make a package, hope the name is free... or we would have to use short eslcets)
| @@ -156,12 +156,19 @@ init([]) -> | |||
| PG = | |||
| {pg, | |||
| {pg, start_link, [mim_scope]}, | |||
| permanent, infinity, supervisor, [pg]}, | |||
| permanent, infinity, worker, [pg]}, | |||
There was a problem hiding this comment.
Good catch, this wasn't a supervisor. But now this and the one below are the only worker whose death timeout is set to infinity while all the others have a brutal_kill. Out of the scope of this PR but we should give it a thought some day (I'm already doing so 😇 )
There was a problem hiding this comment.
rewrite it all, there is a lot of repetition in this module...
mod_muc.erl could be moved in a separate PR into |
|
elasticsearch_and_cassandra_25 / elasticsearch_and_cassandra_mnesia / c5faa49 small_tests_24 / small_tests / c5faa49 small_tests_25_arm64 / small_tests / c5faa49 small_tests_25 / small_tests / c5faa49 ldap_mnesia_24 / ldap_mnesia / c5faa49 ldap_mnesia_25 / ldap_mnesia / c5faa49 dynamic_domains_pgsql_mnesia_24 / pgsql_mnesia / c5faa49 dynamic_domains_mysql_redis_25 / mysql_redis / c5faa49 dynamic_domains_pgsql_mnesia_25 / pgsql_mnesia / c5faa49 dynamic_domains_mssql_mnesia_25 / odbc_mssql_mnesia / c5faa49 pgsql_mnesia_24 / pgsql_mnesia / c5faa49 pgsql_cets_25 / pgsql_cets / c5faa49 mysql_redis_25 / mysql_redis / c5faa49 pgsql_mnesia_25 / pgsql_mnesia / c5faa49 internal_mnesia_25 / internal_mnesia / c5faa49 mssql_mnesia_25 / odbc_mssql_mnesia / c5faa49 |
This PR checks if the CETS code from the feature branch passes the tests.
Ready for the final review and merge.
The following will be done after the merge:
component_backendinternal_databasesand the need to specifymnesiaexplicitly.