Skip to content

fm: make sp_ereport_ingester status output less gaslighty#10382

Open
hawkw wants to merge 3 commits intomainfrom
eliza/stop-ingester-gaslighting
Open

fm: make sp_ereport_ingester status output less gaslighty#10382
hawkw wants to merge 3 commits intomainfrom
eliza/stop-ingester-gaslighting

Conversation

@hawkw
Copy link
Copy Markdown
Member

@hawkw hawkw commented May 6, 2026

Originally, when implementing the sp_ereport_ingester background task, I had the foolish and bad idea to make the task's status output omit entries for SPs which we communicated with successfully but which had no new ereports, thinking that it was "not interesting" to list them. Unfortunately, this also means that the total count of SPs that the task talked to, and the count of total HTTP requests sent, will also not count any SPs which had no ereports. This behavior is incredibly misleading to anyone who isn't aware that they are being intentionally excluded, which, as it turns out, includes me (since I had forgotten that I did this).

This commit fixes this by not going out of our way to exclude these SPs from the status output. It turns out that this actually makes the code somewhat simpler, which is another argument in favor of the idea that I never should have done this. Sigh.

Fixes #10380

Originally, when implementing the `sp_ereport_ingester` background task,
I had the foolish and bad idea to make the task's status output omit
entries for SPs which we communicated with successfully but which had no
new ereports, thinking that it was "not interesting" to list them.
Unfortunately, this also means that the total count of SPs that the task
talked to, and the count of total HTTP requests sent, will also not
count any SPs which had no ereports. This behavior is *incredibly
misleading* to anyone who isn't aware that they are being intentionally
excluded, which, as it turns out, includes me (since I had forgotten
that I did this).

This commit fixes this by not going out of our way to exclude these SPs
from the status output. It turns out that this actually makes the code
somewhat simpler, which is another argument in favor of the idea that I
never should have done this. Sigh.

Fixes #10380
@hawkw hawkw requested review from mergeconflict and smklein May 6, 2026 03:05
@hawkw hawkw added Debugging For when you want better data in debugging an issue (log messages, post mortem debugging, and more) fault-management Everything related to the fault-management initiative (RFD480 and others) labels May 6, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Debugging For when you want better data in debugging an issue (log messages, post mortem debugging, and more) fault-management Everything related to the fault-management initiative (RFD480 and others)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

sp_ereport_ingester's status is misleading

1 participant