sessionctx, variable: add skip name resolve #25464
Conversation
ti-chi-bot
added
the
size/S
morgo
added
sig/sql-infra
| if ip == "127.0.0.1" { | ||
| return []string{variable.DefHostname} | ||
| } | ||
| skipNameResolve, err := s.GetSessionVars().GlobalVarsAccessor.GetGlobalSysVar(variable.SkipNameResolve) | ||
| if err == nil && variable.TiDBOptOn(skipNameResolve) { |
There was a problem hiding this comment.
Are you intended to continue even when err != nil?
There was a problem hiding this comment.
Yes. The option defaults to OFF (and always will). It's only if can be successfully read as ON that we take the short-path and return the IP address. See: https://dev.mysql.com/doc/refman/8.0/en/server-system-variables.html#sysvar_skip_name_resolve
|
[REVIEW NOTIFICATION] This pull request has been approved by:
To complete the pull request process, please ask the reviewers in the list to review by filling The full list of commands accepted by this bot can be found here. Reviewer can indicate their review by submitting an approval review. |
ti-chi-bot
added
status/LGT2
|
/merge |
|
This pull request has been accepted and is ready to merge. Commit hash: a32bcfe
|
ti-chi-bot
added
the
status/can-merge
|
@morgo: Your PR was out of date, I have automatically updated it for you. At the same time I will also trigger all tests for you: /run-all-tests If the CI test fails, you just re-trigger the test that failed and the bot will merge the PR for you after the CI passes. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the ti-community-infra/tichi repository. |
What problem does this PR solve?
Issue Number: close #9778
Problem Summary:
I was investigating linter issues, when I noticed that the silencing of the error in
getHostByIPmight not be completely correct. If the reverse lookup fails (net.LookupAddr) - is the original IP returned? I expect this behavior is undefined, so authentication might fail when it doesn't need to.Then I was wondering why there is no method to skip name resolution like there is in MySQL, since it is a common problem in large scale systems. There has been a feature request @ #9778
In MySQL skip-name-resolve is non-dynamic, but this does not make so much sense for TiDB, so I've changed it to GLOBAL-only scope.
I also don't think it is correct to silently drop errors from name resolution (although the fact that they were previously noisy and this was fixed in #13989 indicates that skip-name resolve as a feature has merit). My resolution was to reintroduce an error message, but switch it to debug level.
What is changed and how it works?
What's Changed:
Support has been added for skip-name-resolve.
Related changes
pingcap/docs/pingcap/docs-cn:Check List
Tests
Side effects
Release note