Skip to content

external mode alerts#1020

Open
b0a7 wants to merge 9 commits intorocket-pool:masterfrom
b0a7:topic/externalModeAlerts
Open

external mode alerts#1020
b0a7 wants to merge 9 commits intorocket-pool:masterfrom
b0a7:topic/externalModeAlerts

Conversation

@b0a7
Copy link

@b0a7 b0a7 commented Feb 28, 2026

This PR limits some of the alerts that appear when your EC/CC are externally managed. Critical warnings about ports that aren't open aren't applicable in an externally managed configuration. For port warnings, the configuration is checked in the code that checks for port connectivity.

Tested via new unit tests

Copy link
Member

@0xfornax 0xfornax left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We either have both clients running locally or externally managed so the code could be simplified and a few tests removed.

@jshufro
Copy link
Contributor

jshufro commented Mar 3, 2026

I actually disagree with the premise here. Low disk space alerts are important, even for VC-only machines. They have a tendency to fill up with logs or docker images- and when they do fill up, clients can't write-ahead to slashing protection databases and stop attesting.

It seems to me that the AlertEnabled_LowDiskSpaceCritical flag can be use by anyone who wants to manually disable these alerts. The default should be on.

However, node_filesystem_avail_bytes is a bad metric to threshold against. The alert should default to comparing to total_bytes so we can fire when below, say, 15% of the drive's total space remains.

@b0a7
Copy link
Author

b0a7 commented Mar 3, 2026

We either have both clients running locally or externally managed so the code could be simplified and a few tests removed.

fixed d6b3398

@b0a7
Copy link
Author

b0a7 commented Mar 3, 2026

I actually disagree with the premise here. Low disk space alerts are important, even for VC-only machines. They have a tendency to fill up with logs or docker images- and when they do fill up, clients can't write-ahead to slashing protection databases and stop attesting.

It seems to me that the AlertEnabled_LowDiskSpaceCritical flag can be use by anyone who wants to manually disable these alerts. The default should be on.

However, node_filesystem_avail_bytes is a bad metric to threshold against. The alert should default to comparing to total_bytes so we can fire when below, say, 15% of the drive's total space remains.

I agree it is a better idea to keep the critical disk space alerts and adjust the threshold. To simplify things I reverted the changes about the disk space alerts and will just focus on the port checks since it is a more obvious problem & solution.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants