Skip to content

Increase server.limits.checkin_limit.max_body_byte_size default#6418

Draft
ebeahan wants to merge 2 commits intoelastic:mainfrom
ebeahan:increase-checkin-body-size-limit-default
Draft

Increase server.limits.checkin_limit.max_body_byte_size default#6418
ebeahan wants to merge 2 commits intoelastic:mainfrom
ebeahan:increase-checkin-body-size-limit-default

Conversation

@ebeahan
Copy link
Member

@ebeahan ebeahan commented Feb 24, 2026

Users of Synthetics have experienced errors when passing a larger number of private location monitors in a single Agent policy. Support and Observability teams asked if we can explore raising the default max_body_byte_size for server.limits.checkin_limit.max_body_byte_size from it's current setting, 1 MB.

Resolves #5549

Relates elastic/elastic-agent#11981

@ebeahan ebeahan self-assigned this Feb 24, 2026
@ebeahan ebeahan added enhancement New feature or request backport-skip Skip notification from the automated backport with mergify labels Feb 24, 2026
@ebeahan
Copy link
Member Author

ebeahan commented Feb 24, 2026

@lucabelluccini from the Support side, do we have any idea of a new default to propose which would address times when we've seen this problem for Synthetics?

In this PR, I've proposed doubling as as a starting point.

@lucabelluccini
Copy link
Contributor

This analysis is purely driven by ECH stats.

I we take the last 7 days of ECH activity, we have ~24.5K errors Bad request: unable to decode checkin request: http: request body too large, affecting ~20 unique deployments. Note this error being logged by Fleet Server has been added only on recent versions.

If we take the ECH activity, check-in requests to Fleet then filter by requests over 4MB returning 400 status code over 1 day, we get:
image

Taking > 1MB returning 400 over 1 day:
image

Keeping aside the weird case of 60MB of check-in size and some 10MB-ish, it looks like 8MB would cover most cases.


Other situations we spotted (not on ECH) required to bump to 8MB or 16MB (~7k monitors including browser ones).

I think the best bump change would be 8MB, but I do not know to which extent this might affect small Integration Servers (1GB).

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

Labels

backport-skip Skip notification from the automated backport with mergify enhancement New feature or request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Revisit default values for max request body size limits.

2 participants