settings,*: give better names to certain settings

[knz]

Fixes #109046.

Now we have a framework that allows us to rename cluster settings
while preserving backward-compatibility, we can clean up the settings
whose name didn't pass the linter, and had been "grand-fathered". This
also simplifies the linter.

The following have been renamed:

Previous name	New name
kv.raft_log.disable_synchronization_unsafe	kv.raft_log.synchronization.disabled
sql.trace.log_statement_execute	sql.log.all_statements.enabled
cloudstorage.gs.chunking.retry_timeout	cloudstorage.gs.chunking.per_chunk_retry.timeout
trace.debug.enable	trace.http_debug_endpoint.enabled
bulkio.restore.memory_monitor_ssts	bulkio.restore.sst_memory_limit.enabled
bulkio.restore.use_simple_import_spans	bulkio.restore.simple_import_spans.enabled
bulkio.backup.export_request_verbose_tracing	bulkio.backup.verbose_tracing.enabled
changefeed.idle_timeout	changefeed.auto_idle.timeout

The following name suffixes have been adjusted to follow our linter
guidelines:

bulkio.backup.split_keys_on_timestamps - added .enabled
changefeed.balance_range_distribution.enable - renamed to .enabled
changefeed.batch_reduction_retry_enabled - renamed to .enabled
changefeed.new_pubsub_sink_enabled - renamed to .enabled
changefeed.new_webhook_sink_enabled - renamed to .enabled
changefeed.permissions.require_external_connection_sink - added .enabled
debug.panic_on_failed_assertions - added .enabled
diagnostics.reporting.send_crash_reports - added .enabled
kv.closed_timestamp.follower_reads_enabled - renamed to .enabled
kv.range_merge.queue_enabled - renamed to .enabled
kv.range_split.by_load_enabled - renamed to .enabled
kv.transaction.parallel_commits_enabled - renamed to .enabled
kv.transaction.write_pipelining_enabled - renamed to .enabled
kv.transaction.write_pipelining_max_batch_size - renamed to .max_batch_size
server.clock.forward_jump_check_enabled - renamed to .enabled
server.oidc_authentication.autologin - added .enabled
server.web_session_timeout - renamed to .timeout
sql.distsql.flow_stream_timeout - renamed to .timeout
sql.metrics.statement_details.dump_to_logs - added .enabled
stream_replication.job_liveness_timeout - renamed to .timeout

Release note (cli change): The following user-visible cluster settings
have been renamed. The previous name is still available for backward
compatibility.

Previous name	New name
server.web_session_timeout	server.web_session.timeout
kv.closed_timestamp.follower_reads_enabled	kv.closed_timestamp.follower_reads.enabled
kv.range_split.by_load_enabled	kv.range_split.by_load.enabled
changefeed.balance_range_distribution.enable	changefeed.balance_range_distribution.enabled
changefeed.batch_reduction_retry_enabled	changefeed.batch_reduction_retry.enabled
server.clock.forward_jump_check_enabled	server.clock.forward_jump_check.enabled
server.oidc_authentication.autologin	server.oidc_authentication.autologin.enabled
sql.metrics.statement_details.dump_to_logs	sql.metrics.statement_details.dump_to_logs.enabled
sql.trace.log_statement_execute	sql.log.all_statements.enabled
trace.debug.enable	trace.http_debug_endpoint.enabled

Epic: CRDB-28893

[cockroach-teamcity]

This change is 

[stevendanna]

@dt Now that we have the option to rename things, I wonder if we want to make a follow-up pass to @knz's work here to address some of our long-standing setting naming gripes?

[j82w]

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @ericharmeling, @knz, @samiskin, and @yuzefovich)

---

-- commits line 4 at r3:
I created this issue about possibly improving the naming. What do you think? #108508

---

-- commits line 10 at r3:
Shouldn't there be a release note that cluster settings are getting aliased with new names?

[knz]

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @ericharmeling, @j82w, @samiskin, and @yuzefovich)

---

-- commits line 4 at r3:

Previously, j82w (Jake) wrote…

I created this issue about possibly improving the naming. What do you think? #108508

Yes I saw. I started this work in light of your issue.
However, note that the concern you raised there already has a plan under epic CRDB-6671. There's multiple steps there before we can consider directly what you suggest.

---

-- commits line 10 at r3:

Previously, j82w (Jake) wrote…

Shouldn't there be a release note that cluster settings are getting aliased with new names?

I thought I included this release note already. Doesn't it show up for you?

[j82w]

-- commits line 10 at r3:

Previously, knz (Raphael 'kena' Poss) wrote…

I thought I included this release note already. Doesn't it show up for you?

The 2nd commit has the release note. I left the comment looking at the first commit and forgot to delete it.

[j82w]

-- commits line 4 at r3:

Previously, knz (Raphael 'kena' Poss) wrote…

Yes I saw. I started this work in light of your issue.
However, note that the concern you raised there already has a plan under epic CRDB-6671. There's multiple steps there before we can consider directly what you suggest.

If it already supports alias and a rename is being done why can't we incorporate the setting type into the name now?

[knz]

If it already supports alias and a rename is being done why can't we incorporate the setting type into the name now?

Because the recommendation you made is only one of the ways to solve the problem that you identified. We are going to explore other ways to solve it first.

[ericharmeling]

Great usability improvement. We'll definitely want docs follow-up on this.

@florence-crl

[knz]

@stevendanna has communicated in private his lack of further objection on this one.

TFYRs

bors r=stevendanna,ericharmeling

[craig]

Build failed (retrying...):

• Bazel Essential CI (Cockroach)

[craig]

Build succeeded:

• Bazel Essential CI (Cockroach)