chore: update atlas config

[lgalabru]

In this PR, we're bumping max_uninstantiated_attachments to 50k and uninstantiated_attachments_expire_after to 1 day.

[jcnelson]

LGTM. @CharlieC3 @deantchi when deploying, can you keep some nodes with the old setting, just in case this puts too much load on the p2p threads of the new nodes?

[CharlieC3]

@jcnelson Are these settings configurable in the config.toml? It would be a big help if they are.
cc @lgalabru

[codecov]

Codecov Report

Merging #3612 (70e0483) into develop (04a504b) will increase coverage by 1.68%.
The diff coverage is n/a.

@@             Coverage Diff             @@
##           develop    #3612      +/-   ##
===========================================
+ Coverage    83.80%   85.49%   +1.68%     
===========================================
  Files          297      298       +1     
  Lines       276103   276674     +571     
===========================================
+ Hits        231397   236550    +5153     
+ Misses       44706    40124    -4582     

Impacted Files	Coverage Δ
src/net/atlas/mod.rs	74.11% <ø> (ø)

... and 55 files with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[lgalabru]

I agree that ideally, this settings would be exposed in the config.
Could someone working full-time on the blockchain team could follow up?
It sounds like a good first issue (cc @igorsyl / @kantai / @jcnelson for dispatch).

[pavitthrap]

From blockchain meeting: test this live on mainnet for a week

[kantai]

Now that we have a configurable version of this PR (#3618), should we close this one?

[jbencin]

Now that we have a configurable version of this PR (#3618), should we close this one?

PR #3618 is still using the old default values. If the testing by @CharlieC3 and @zone117x show these values to be an improvement, we can merge this PR to update the defaults

[pavitthrap]

Blockchain meeting: testing halfway complete, need to see how these values affect attachment performance