mirrors/gotosocial
1
0
Fork 0
mirror of https://codeberg.org/superseriousbusiness/gotosocial.git synced 2024-12-19 07:31:53 +03:00
Code Issues Projects Releases Packages Wiki Activity
gotosocial/vendor/github.com/KimMachineGun/automemlimit/README.md

49 lines
1.6 KiB
Markdown
Raw Normal View History

[feature] Inherit resource limits from cgroups (#1336) When GTS is running in a container runtime which has configured CPU or memory limits or under an init system that uses cgroups to impose CPU and memory limits the values the Go runtime sees for GOMAXPROCS and GOMEMLIMIT are still based on the host resources, not the cgroup. At least for the throttling middlewares which use GOMAXPROCS to configure their queue size, this can result in GTS running with values too big compared to the resources that will actuall be available to it. This introduces 2 dependencies which can pick up resource contraints from the current cgroup and tune the Go runtime accordingly. This should result in the different queues being appropriately sized and in general more predictable performance. These dependencies are a no-op on non-Linux systems or if running in a cgroup that doesn't set a limit on CPU or memory. The automatic tuning of GOMEMLIMIT can be disabled by either explicitly setting GOMEMLIMIT yourself or by setting AUTOMEMLIMIT=off. The automatic tuning of GOMAXPROCS can similarly be counteracted by setting GOMAXPROCS yourself.
2023-01-17 23:59:04 +03:00
# automemlimit
[![Go Reference](https://pkg.go.dev/badge/github.com/KimMachineGun/automemlimit.svg)](https://pkg.go.dev/github.com/KimMachineGun/automemlimit)
[![Go Report Card](https://goreportcard.com/badge/github.com/KimMachineGun/automemlimit)](https://goreportcard.com/report/github.com/KimMachineGun/automemlimit)
[![Test](https://github.com/KimMachineGun/automemlimit/actions/workflows/test.yml/badge.svg?branch=main)](https://github.com/KimMachineGun/automemlimit/actions/workflows/test.yml)
Automatically set `GOMEMLIMIT` to match Linux [cgroups(7)](https://man7.org/linux/man-pages/man7/cgroups.7.html) memory limit.
See more details about `GOMEMLIMIT` [here](https://tip.golang.org/doc/gc-guide#Memory_limit).
## Installation
```shell
go get github.com/KimMachineGun/automemlimit@latest
```
## Usage
```go
package main
// By default, it sets `GOMEMLIMIT` to 90% of cgroup's memory limit.
// You can find more details of its behavior from the doc comment of memlimit.SetGoMemLimitWithEnv.
import _ "github.com/KimMachineGun/automemlimit"
```
or
```go
package main
import "github.com/KimMachineGun/automemlimit/memlimit"
func init() {
[chore]: Bump github.com/KimMachineGun/automemlimit from 0.2.6 to 0.3.0 (#2165)
2023-08-28 09:59:08 +03:00
memlimit.SetGoMemLimitWithOpts(
memlimit.WithRatio(0.9),
memlimit.WithEnv(),
memlimit.WithProvider(memlimit.FromCgroup),
)
[feature] Inherit resource limits from cgroups (#1336) When GTS is running in a container runtime which has configured CPU or memory limits or under an init system that uses cgroups to impose CPU and memory limits the values the Go runtime sees for GOMAXPROCS and GOMEMLIMIT are still based on the host resources, not the cgroup. At least for the throttling middlewares which use GOMAXPROCS to configure their queue size, this can result in GTS running with values too big compared to the resources that will actuall be available to it. This introduces 2 dependencies which can pick up resource contraints from the current cgroup and tune the Go runtime accordingly. This should result in the different queues being appropriately sized and in general more predictable performance. These dependencies are a no-op on non-Linux systems or if running in a cgroup that doesn't set a limit on CPU or memory. The automatic tuning of GOMEMLIMIT can be disabled by either explicitly setting GOMEMLIMIT yourself or by setting AUTOMEMLIMIT=off. The automatic tuning of GOMAXPROCS can similarly be counteracted by setting GOMAXPROCS yourself.
2023-01-17 23:59:04 +03:00
memlimit.SetGoMemLimitWithEnv()
memlimit.SetGoMemLimit(0.9)
memlimit.SetGoMemLimitWithProvider(memlimit.Limit(1024*1024), 0.9)
memlimit.SetGoMemLimitWithProvider(memlimit.FromCgroup, 0.9)
memlimit.SetGoMemLimitWithProvider(memlimit.FromCgroupV1, 0.9)
[chore]: Bump github.com/KimMachineGun/automemlimit from 0.3.0 to 0.4.0 (#2440) Bumps [github.com/KimMachineGun/automemlimit](https://github.com/KimMachineGun/automemlimit) from 0.3.0 to 0.4.0. - [Commits](https://github.com/KimMachineGun/automemlimit/compare/v0.3.0...v0.4.0) --- updated-dependencies: - dependency-name: github.com/KimMachineGun/automemlimit dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2023-12-11 13:35:15 +03:00
memlimit.SetGoMemLimitWithProvider(memlimit.FromCgroupHybrid, 0.9)
[feature] Inherit resource limits from cgroups (#1336) When GTS is running in a container runtime which has configured CPU or memory limits or under an init system that uses cgroups to impose CPU and memory limits the values the Go runtime sees for GOMAXPROCS and GOMEMLIMIT are still based on the host resources, not the cgroup. At least for the throttling middlewares which use GOMAXPROCS to configure their queue size, this can result in GTS running with values too big compared to the resources that will actuall be available to it. This introduces 2 dependencies which can pick up resource contraints from the current cgroup and tune the Go runtime accordingly. This should result in the different queues being appropriately sized and in general more predictable performance. These dependencies are a no-op on non-Linux systems or if running in a cgroup that doesn't set a limit on CPU or memory. The automatic tuning of GOMEMLIMIT can be disabled by either explicitly setting GOMEMLIMIT yourself or by setting AUTOMEMLIMIT=off. The automatic tuning of GOMAXPROCS can similarly be counteracted by setting GOMAXPROCS yourself.
2023-01-17 23:59:04 +03:00
memlimit.SetGoMemLimitWithProvider(memlimit.FromCgroupV2, 0.9)
}
```
Reference in a new issue Copy permalink