Added support to run etcd client calls only when in multi-node case #504
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
/work-in-progress |
|
@aaronfern Command |
efbd42f to
bb64b16
Compare
bb64b16 to
bbf1ffa
Compare
bbf1ffa to
7d3be46
Compare
timuthy
suggested changes
Jul 14, 2022
pkg/miscellaneous/miscellaneous.go
Outdated
Comment on lines
345
to
349
| etcdConfigForTest := os.Getenv("ETCD_CONF") | ||
| if etcdConfigForTest != "" { | ||
| return false | ||
| } | ||
| inputFileName = "/var/etcd/config/etcd.conf.yaml" |
There was a problem hiding this comment.
This should be transferred into a utility function. I also see this code being duplicated multiple times in the code base.
There was a problem hiding this comment.
Can you please also make other areas in the code use this utility?
timuthy
reviewed
Jul 14, 2022
pkg/miscellaneous/miscellaneous.go
Outdated
| @@ -338,6 +338,36 @@ func ProbeEtcd(ctx context.Context, clientFactory etcdClient.Factory, logger *lo | |||
| return nil | |||
| } | |||
|
|
|||
| // IsMultiNode determines whether a pod is part of a multi node setup or not | |||
| // This is determined by checking the `initial-cluster` of the etcd configmap to check the number of members expected. | |||
| func IsMultiNode() bool { | |||
There was a problem hiding this comment.
Maybe we can align this more with that is already used here
to not have logic duplication all over the place.7d3be46 to
4e2f97a
Compare
fa7fe27 to
c60a1d7
Compare
c60a1d7 to
049bfa5
Compare
timuthy
suggested changes
Jul 18, 2022
| if etcdConfigForTest != "" { | ||
|
|
||
| inputFileName = GetConfigFilePath() | ||
| if inputFileName != EtcdConfigFilePath { |
There was a problem hiding this comment.
Can you explain why this check is required or was not there before?
There was a problem hiding this comment.
So ideally if we are in test-mode (ETCD_CONF is set) then we return an empty service endpoint so that the default one is used (http://127.0.0.1:2379)
We send a proper service endpoint only when running in a cluster
There was a problem hiding this comment.
Hm, I can't really judge this case but it seems very implicit and also confusing (because why should we then supprt setting a config via ETCD_CONF ?!). At a minimum a comment would be nice why such a check is performed.
There was a problem hiding this comment.
Yes, I understand, and we will be picking this up and finding a better way to implement this while refactoring the code base
For the time being I have added a comment here
pkg/server/backuprestoreserver.go
Outdated
| err := m.PromoteMember(ctx) | ||
| if err == nil { | ||
| break | ||
| if miscellaneous.IsMultiNode(b.logger) { |
There was a problem hiding this comment.
Should we rather use the restoreOpts.OriginalClusterSize here, so that we don't have to decode the configYML again?
There was a problem hiding this comment.
Hmm, you're right, that makes sense
Will make that change
de812f3 to
aaeeada
Compare
aaeeada to
8dbeddc
Compare
8dbeddc to
d7b9ac9
Compare
aaronfern
added a commit
to aaronfern/etcd-backup-restore
that referenced
this pull request
Jul 21, 2022
…ardener#504) * Added support to run etcd client calls only when in multi-node case * Addressed review comments * Used values from restoreOpts rather than another func call
aaronfern
added a commit
that referenced
this pull request
Jul 21, 2022
* Added support to run etcd client calls only when in multi-node case (#504) * Added support to run etcd client calls only when in multi-node case * Addressed review comments * Used values from restoreOpts rather than another func call * Temp fix: skip single member restoration in case of data-dir invalid. (#501) * Temp fix: skip single member restoration in case of data-dir invalid. * Small fixes. * Improved error log. * Fixed a bug in scaleup feature which present in func: IsMemberInCluster. * Fix the unit tests. * Address the review comments. * Handle the case of when storageProvider is not configured. * Address the review comments. * Address the review comments. * Assigned the correct Peer Address to Etcd after it restores from the backup-bucket. (#505) * Assigned the correct Peer address to the Etcd after restoration. * Fix the unit tests and some refactoring. * Improve some logs. * Address the review comments. * To update the peer address of existing cluster. * Don't update peer URL when promoting member (#506) * Don't update peer URL when promoting member * Update interface name in member package Co-authored-by: Ishan Tyagi <42602577+ishan16696@users.noreply.github.com>
9 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What this PR does / why we need it:
This PR adds support to allow etcd calls only if
backup-restoredetermines that it is in a multi-node setup.This removes dependencies in a single node case of requiring network policies for etcd calls to succeed
Which issue(s) this PR fixes:
Fixes #
Special notes for your reviewer:
Release note: