[
https://issues.apache.org/jira/browse/ZOOKEEPER-4557?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17554576#comment-17554576
]
Christopher Tubbs commented on ZOOKEEPER-4557:
----------------------------------------------
[[email protected]] It is related to SLF4J in the sense that ZooKeeper is using the
SLF4J MDC API:
{code}
[christopher@localhost zookeeper] ((release-3.8.0-1))$ grep MDC **/*.java
zookeeper-server/src/main/java/org/apache/zookeeper/ClientCnxn.java:import
org.slf4j.MDC;
zookeeper-server/src/main/java/org/apache/zookeeper/ClientCnxn.java:
MDC.put("myid", hostPort);
zookeeper-server/src/main/java/org/apache/zookeeper/server/quorum/QuorumPeerConfig.java:import
org.slf4j.MDC;
zookeeper-server/src/main/java/org/apache/zookeeper/server/quorum/QuorumPeerConfig.java:
MDC.put("myid", myIdString);
{code}
It is true that the change in behavior is not SLF4J's fault. But, my comment
wasn't blaming SLF4J. I was merely pointing out that ZooKeeper cannot rely on
the MDC feature, because it's not universally supported by all SLF4J runtime
implementation libraries. So, my wording was imprecise. It's not that the
feature is fragile in SLF4J. What is fragile is ZooKeeper's reliance on it
being implemented in the logging runtime implementation.
> MDC for myid is missing in children threads due to log4j vs logback
> incompatibility
> -----------------------------------------------------------------------------------
>
> Key: ZOOKEEPER-4557
> URL: https://issues.apache.org/jira/browse/ZOOKEEPER-4557
> Project: ZooKeeper
> Issue Type: Bug
> Affects Versions: 3.8.0
> Reporter: Shinya Yoshida
> Priority: Major
>
> In 3.8.0, zookeeper migrated from log4j to logback in ZOOKEEPER-4427.
> Unfortunately, logback doesn't use InheritableThreadLocal for MDC
> implementation because it's too dangerous unlike log4j does.
> https://jira.qos.ch/browse/LOGBACK-422
> https://jira.qos.ch/browse/LOGBACK-624
> This causes incompatibility of logging behavior.
> Now we missed pretty myid MDC in the logs.
> You can easily reproduce this issue by just downloading the zookeeper
> artifact and start-up locally.
> Zookeeper 3.7.1:
> {code:java}
> 2022-06-14 15:28:04,279 [myid:1] - INFO
> [ListenerHandler-localhost/127.0.0.1:3888:QuorumCnxManager$Listener$ListenerHandler@1071]
> - 1 is accepting connections now, my election bind port:
> localhost/127.0.0.1:3888
> 2022-06-14 15:28:04,280 [myid:1] - INFO [main:ZKAuditProvider@42] -
> ZooKeeper audit is disabled.
> 2022-06-14 15:28:04,284 [myid:1] - INFO
> [QuorumPeer[myid=1](plain=[0:0:0:0:0:0:0:0]:2181)(secure=disabled):QuorumPeer@1438]
> - LOOKING
> 2022-06-14 15:28:04,284 [myid:1] - INFO
> [QuorumPeer[myid=1](plain=[0:0:0:0:0:0:0:0]:2181)(secure=disabled):FastLeaderElection@945]
> - New election. My id = 1, proposed zxid=0x0
> 2022-06-14 15:28:04,290 [myid:1] - INFO
> [WorkerReceiver[myid=1]:FastLeaderElection$Messenger$WorkerReceiver@390] -
> Notification: my state:LOOKING; n.sid:1, n.state:LOOKING, n.leader:1,
> n.round:0x1, n.peerEpoch:0x0, n.zxid:0x0, message format version:0x2,
> n.config version:0x0
> 2022-06-14 15:28:04,292 [myid:1] - WARN
> [QuorumConnectionThread-[myid=1]-2:QuorumCnxManager@401] - Cannot open
> channel to 3 at election address localhost/127.0.0.1:3890
> {code}
> zookeeper 3.8.0
> {code:java}
> 2022-06-14 15:28:51,461 [myid:] - INFO
> [ListenerHandler-localhost/127.0.0.1:3888:o.a.z.s.q.QuorumCnxManager$Listener$ListenerHandler@1071]
> - 1 is accepting connections now, my election bind port:
> localhost/127.0.0.1:3888
> 2022-06-14 15:28:51,463 [myid:1] - INFO [main:o.a.z.a.ZKAuditProvider@42] -
> ZooKeeper audit is disabled.
> 2022-06-14 15:28:51,467 [myid:] - INFO
> [QuorumPeer[myid=1](plain=[0:0:0:0:0:0:0:0]:2181)(secure=disabled):o.a.z.s.q.QuorumPeer@1455]
> - LOOKING
> 2022-06-14 15:28:51,467 [myid:] - INFO
> [QuorumPeer[myid=1](plain=[0:0:0:0:0:0:0:0]:2181)(secure=disabled):o.a.z.s.q.FastLeaderElection@946]
> - New election. My id = 1, proposed zxid=0x0
> 2022-06-14 15:28:51,469 [myid:] - INFO
> [WorkerReceiver[myid=1]:o.a.z.s.q.FastLeaderElection$Messenger$WorkerReceiver@391]
> - Notification: my state:LOOKING; n.sid:1, n.state:LOOKING, n.leader:1,
> n.round:0x1, n.peerEpoch:0x0, n.zxid:0x0, message format version:0x2,
> n.config version:0x0
> 2022-06-14 15:28:51,475 [myid:] - WARN
> [QuorumConnectionThread-[myid=1]-1:o.a.z.s.q.QuorumCnxManager@401] - Cannot
> open channel to 2 at election address localhost/127.0.0.1:3889
> {code}
> Now we couldn't see myid in the most of logs.
> We need to propagate myid MDC to other threads in the code.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
