[Website]: Cluster formation deployment type

[stefnestor]

Type of issue

None

What documentation page is affected

https://www.elastic.co/docs/deploy-manage/distributed-architecture/discovery-cluster-formation

What happened?

👋 howdy!

Couple points of commentary

• (A) I'm not sure why this guide doesn't tag as applying across deployment types since call out agrees it does.

  applies_to:
    deployment:
      ess: all
      ece: all
      eck: all
      self: all
    serverless: all

• (B) I agree with this caveat for Serverless, but FWIW while ECH automates/handles this for admins they do still need to know about and worry about this section especially "cluster fault detection" so it is not just FYI vs is directly needed for troubleshooting. For example on ECH, user can API restart nodes & needs to know not to interrupt/mess up this automation.
• (C) baby typo

Additional info

No response

[shainaraskas]

stack covers all of those deployment types (ess/ece/eck/self)

I'd argue that we should take serverless out of the note because I don't know if it even works the same way under the hood (therefore not needing a serverless tag on the page)

can you give me a link to where the ech-relevant procedures are? I don't see anything that should be used for more than informational purposes at a quick glance.

[stefnestor]

can you give me a link to where the ech-relevant procedures are? I don't see anything that should be used for more than informational purposes at a quick glance.

IMO my earlier "For example on ECH, user can API restart nodes & needs to know not to interrupt/mess up this automation." disagrees with this sentiment. Support has users who interrupt quorum-based decision making & who have to troubleshoot lagging cluster state.

stack covers all of those deployment types (ess/ece/eck/self)

It might be a usability/standaridzation issue then, sorry 🙃, but that's not my interpretation. Ex: Cluster Fault Detection is a varying word repeat of Troubleshoot an unstable cluster which does have all its tags

[shainaraskas]

hey stef, sorry, just circling back

IMO my earlier "For example on ECH, user can API restart nodes & needs to know not to interrupt/mess up this automation." disagrees with this sentiment. Support has users who interrupt quorum-based decision making & who have to troubleshoot lagging cluster state.

to your point on item 1: while the information is relevant, the configuration settings are not. that's the point of that note. editing the note a bit to see if I can land the plane

It might be a usability/standaridzation issue then, sorry 🙃, but that's not my interpretation. Ex: Cluster Fault Detection is a varying word repeat of Troubleshoot an unstable cluster which does have all its tags

the troubleshooting pages are mistagged. I have to fix them at some point soon