Community Contribution License

All community contributions in this pull request are licensed to the project maintainers
under the terms of theApache 2 license.
By creating this pull request I represent that I have the right to license the
contributions to the project maintainers under the Apache 2 license.

Description

This commit improves the performance of xlStorage.openFile and thus AppendFile by ~33% when the file's parent directory already exists. The performance improvement comes from changing openFile to optimistically attempt to open the file and only create the parent directories if os.OpenFile fails with ErrNotExist.

The performance impact of a failed call to os.OpenFile when the parent directory does not exist and must be created is about ~300ns on Linux, which is far less than the benefit this change provides for the likely more common case of the parent directory existing.

Benchmark results:

goos: linuxgoarch: arm64pkg: github.com/minio/minio/cmd                      │ base.10.txt │             new.10.txt              │                      │   sec/op    │   sec/op     vs base                │XLStorageAppendFile-4   9.023µ ± 1%   5.961µ ± 0%  -33.94% (p=0.000 n=10)                      │ base.10.txt  │              new.10.txt              │                      │     B/op     │     B/op      vs base                │XLStorageAppendFile-4   2.281Ki ± 1%   1.672Ki ± 0%  -26.71% (p=0.000 n=10)                      │ base.10.txt │             new.10.txt             │                      │  allocs/op  │ allocs/op   vs base                │XLStorageAppendFile-4    50.00 ± 0%   35.00 ± 0%  -30.00% (p=0.000 n=10)

goos: linuxgoarch: amd64pkg: github.com/minio/minio/cmdcpu: AMD EPYC-Milan Processor                      │ base.10.txt │             new.10.txt              │                      │   sec/op    │   sec/op     vs base                │XLStorageAppendFile-4   23.07µ ± 3%   16.10µ ± 3%  -30.21% (p=0.000 n=10)                      │ base.10.txt  │              new.10.txt              │                      │     B/op     │     B/op      vs base                │XLStorageAppendFile-4   2.281Ki ± 1%   1.672Ki ± 0%  -26.71% (p=0.000 n=10)                      │ base.10.txt │             new.10.txt             │                      │  allocs/op  │ allocs/op   vs base                │XLStorageAppendFile-4    50.00 ± 0%   35.00 ± 0%  -30.00% (p=0.000 n=10)

Motivation and Context

Motivation is simply performance. As for context, I noticed this while reviewing the minio code.

How to test this PR?

The existing tests cover this logic.

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Optimization (provides speedup with no functional changes)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• Fixes a regression (If yes, please addcommit-id orPR # here)
• Unit tests added/updated
• Internal documentation updated
• Create a documentation update requesthere