-
Notifications
You must be signed in to change notification settings - Fork 32
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Bug 1908707: fluentd fails to deliver message with Server returned nothing #73
Conversation
@syedriko: This pull request references Bugzilla bug 1908707, which is valid. The bug has been moved to the POST state. The bug has been updated to refer to the pull request using the external bug tracker. 3 validation(s) were run on this bug
In response to this:
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 kubernetes/test-infra repository. |
/approve |
/test cluster-logging-operator-e2e |
…thing Corrected syntax of flags declarations, added unit tests for the new flags.
/test cluster-logging-operator-e2e |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Few nits that I am not firm on the necessity to resolve in the PR. I leave it to you , but otherwise lgtm
AuthBackEndRoles: map[string]BackendRoleConfig{}, | ||
AuthWhiteListedNames: []string{}, | ||
AuthAdminRole: "", | ||
HTTPReadTimeout: time.Duration(1) * time.Minute, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
For consistency, should we choose the same unit and apply it to all config?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think this dovetails with the next comment. For consistency we'd need to go with the smallest unit we're likely to encounter, which is millisecond, and express a minute as time.Duration(60000) * time.Millisecond. This is arguably more difficult to read than time.Duration(1) * time.Minute. I'm leaning towards the easiest to read - the smallest value with the right units factor.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Per my other comment, as long as the CLI allows us to specify a duration I am satisfied leaving these as-is
@@ -37,5 +37,18 @@ func newFlagSet() *flag.FlagSet { | |||
flagSet.String("auth-admin-role", "", "The name of the only role that will be passed on the request if it is found in the list of roles") | |||
flagSet.String("auth-default-role", "", "The role given to every request unless it has the auth-admin-role") | |||
|
|||
//net/http.Server timeouts for the server side of the proxy | |||
flagSet.Duration("http-read-timeout", time.Duration(1)*time.Minute, "The maximum duration for reading the entire HTTP request. Zero means no timeout.") |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Should we expand the flags to identify the unit? e.g http-read-timeout-sec
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That was my first impulse, too, but I don't think it's the right thing. A flag of type time.Duration accepts many different time units, from nanoseconds to hours. See https://golang.org/pkg/flag/#Duration. Under the covers, https://golang.org/pkg/time/#ParseDuration does the parsing. I had fun with it at https://github.com/syedriko/elasticsearch-proxy/blob/f16618abbf400e2f3536c257df9bcc2cc52e86cf/pkg/config/options_test.go#L320
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can i define a proper duration in configuration on the CLI? It will let me use say 1s
or 10m
? If so, I'm good as-is
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It will even let you define --http-expect-continue-timeout=1h2m3s4ms5us6ns
if you so desire.
…thing Code review feedback
/test cluster-logging-operator-e2e |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
/lgtm
/retest
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: jcantrill, syedriko The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
/cherrypick release-4.6 |
@jcantrill: once the present PR merges, I will cherry-pick it on top of release-4.6 in a new PR and assign it to you. In response to this:
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 kubernetes/test-infra repository. |
@syedriko: All pull requests linked via external trackers have merged: Bugzilla bug 1908707 has been moved to the MODIFIED state. In response to this:
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 kubernetes/test-infra repository. |
@jcantrill: #73 failed to apply on top of branch "release-4.6":
In response to this:
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 kubernetes/test-infra repository. |
Bug 1924258: fluentd fails to deliver message with Server returned nothing
Bug 1924258: fluentd fails to deliver message with Server returned nothing
Bug 1924329: fluentd fails to deliver message with Server returned nothing
Description
Fix for https://bugzilla.redhat.com/show_bug.cgi?id=1908707.
Make net/http transport timeouts and limits configurable. Increase the default HTTP server read/write/idle timeouts from 5 seconds to 1 minute.
/cc @jcantrill
/assign @ewolinetz