=cluster handle aggressively rejoining/replacing nodes from same host/port pair #1083
Conversation
This file contains 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
…pected crash in the future
ktoso
changed the title
ktoso
commented
Nov 1, 2022
| actorIdentifier = "[\(id)]" | ||
| _ = metadata.removeValue(forKey: "actor/path") // discard it | ||
| } else if let path = metadata.removeValue(forKey: "actor/path") { | ||
| actorIdentifier = "[\(path)]" |
There was a problem hiding this comment.
revamping this now that we have more distributed actor types and less paths 👍
ktoso
commented
Nov 1, 2022
| @@ -106,7 +106,7 @@ struct SamplePrettyLogHandler: LogHandler { | |||
|
|
|||
| let file = file.split(separator: "/").last ?? "" | |||
| let line = line | |||
| print("\(self.timestamp()) [\(file):\(line)] [\(nodeInfo)\(Self.CONSOLE_BOLD)\(label)\(Self.CONSOLE_RESET)] [\(level)] \(message)\(metadataString)") | |||
| print("\(self.timestamp()) \(level) [\(nodeInfo)\(Self.CONSOLE_BOLD)\(label)\(Self.CONSOLE_RESET)][\(file):\(line)] \(message)\(metadataString)") | |||
There was a problem hiding this comment.
same formatting as our example formatters in samples
ktoso
commented
Nov 1, 2022
| .lazy | ||
| .filter { member in member.status <= status } | ||
| .count | ||
| } |
There was a problem hiding this comment.
used in test but generally useful
ktoso
commented
Nov 1, 2022
| @@ -223,6 +238,21 @@ extension Cluster { | |||
| } | |||
| } | |||
|
|
|||
| extension Cluster.Membership: Sequence { | |||
There was a problem hiding this comment.
generally useful to for member in membership 👍
ktoso
commented
Nov 1, 2022
| change.member.node.asClusterNode == self.id.node | ||
| { | ||
| return | ||
| } |
There was a problem hiding this comment.
The fix™
Though this deserves a deeper dig into SWIM as well.
Also, SwiftFormat's formatting of multi-line if is annoying omg!
yim-lee
approved these changes
Nov 1, 2022
|
SWIM follow up apple/swift-cluster-membership#91 |
Co-authored-by: Yim Lee <[email protected]>
|
How I love a trailing space being the only failure :P |
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.
Resolves : #1082
Short version: there is an edge case that was not handled well when very aggressively rejoining the cluster from the same host/port.
This can happen in k8s when a pod gets aggressively restarted, or on command line apps when someone joins a cluster "sends just one request and kills the app" since they both then still may be present in SWIM gossip (correctly) as dead, and the cluster may misinterpret this about information about "itself" when the new node joins.
More analysis:
steps:
2022-11-01T13:08:56+0900 trace Client : actor/id=/user/swim actor/path=/user/swim cluster/node=sact://[email protected]:7337 swim/incarnation=0 swim/members/all=["SWIM.Member(SWIMActor(id:sact://RemoteCluster:[email protected]:8337/user/swim, node:sact://RemoteCluster:[email protected]:8337, alive(incarnation: 0), protocolPeriod: 1)", "SWIM.Member(SWIMActor(id:/user/swim, node:sact://[email protected]:7337, alive(incarnation: 0), protocolPeriod: 0)"] swim/members/count=2 swim/ping/origin=sact://RemoteCluster:[email protected]:8337/user/swim swim/ping/payload=membership([SWIM.Member(SWIMActor(id:sact://[email protected]:7337/user/swim, node:sact://[email protected]:7337, suspect(incarnation: 0, suspectedBy: Set([sact://[email protected]:8337#7602583950674506995])), protocolPeriod: 56), SWIM.Member(SWIMActor(id:sact://RemoteCluster:[email protected]:8337/user/swim, node:sact://RemoteCluster:[email protected]:8337, alive(incarnation: 0), protocolPeriod: 0), SWIM.Member(SWIMActor(id:/user/swim, node:sact://[email protected]:7337, alive(incarnation: 0), protocolPeriod: 56)]) swim/ping/seqNr=4 swim/protocolPeriod=1 swim/suspects/count=0 swim/timeoutSuspectsBeforePeriodMax=11 swim/timeoutSuspectsBeforePeriodMin=4 [DistributedCluster] Received ping@4swim/ping/payload=membership([SWIM.Member(SWIMActor(id:sact://[email protected]:7337/user/swim, node:sact://[email protected]:7337, suspect(incarnation: 0, suspectedBy: Set([sact://[email protected]:8337#7602583950674506995])), protocolPeriod: 56),SWIM.Member(SWIMActor(id:sact://[email protected]:7337/user/swim, node:sact://[email protected]:7337, alive(incarnation: 0), protocolPeriod: 56),SWIM.Member(SWIMActor(id:sact://RemoteCluster:[email protected]:8337/user/swim, node:sact://RemoteCluster:[email protected]:8337, alive(incarnation: 0), protocolPeriod: 0)confirmDeadyet -- THIS IS OK. But continuing to act on the removed node's information is NOT ok.Long story short: SWIM tells us that a node on this address was dead, but we know we are not dead -- this should only happen on high level gossip, when we see a .down somewhere about us. So we can ignore this from the SWIM level.
This should also get fixed in SWIM itself though, I'll follow up there.