Skip to content

HEALTHCHECK_START_PERIOD may be too short for slower systems #4231

@arsalanyavari

Description

@arsalanyavari

Problem Statement

The default value for HEALTHCHECK_START_PERIOD seems too short on some lower-resource or slower systems.
In my case, during docker compose up, the web container became unhealthy multiple times before it had enough time to fully start, and as a result relay stayed in created.
My system resources:
RAM: 16Gi
SWAP: 16Gi
DISK: SSD
DIRSK FREE: 86G
CPUS: 4
CPU: Intel Xeon Processor (Skylake, IBRS)
Before increasing this variable, I was seeing:

sentry-self-hosted-web-1   sentry-self-hosted-local   "/etc/sentry/entrypo…"   web   7 minutes ago   Up 7 minutes (unhealthy)   127.0.0.1:9000->9000/tcp

after that the relay service has beed created and do not change...

--
After increasing HEALTHCHECK_START_PERIOD, the issue was resolved.

Solution Brainstorm

I suggest increasing the default value of this variable. If you prefer not to change the default, it would be helpful to mention this in the documentation for users running on slower systems.

Metadata

Metadata

Assignees

No one assigned
    No fields configured for issues without a type.

    Projects

    Status

    No status

    Status

    No status

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions