MINOR: Add ZK migration docs to the packaged docs#13257
Conversation
|
|
||
| <pre> | ||
| Completed migration of metadata from Zookeeper to KRaft | ||
| </pre> |
There was a problem hiding this comment.
nit: this adds an extra newline
| # Sample KRaft cluster controller.properties listening on 9093 | ||
| process.roles=controller | ||
| node.id=3000 | ||
| controller.quorum.voters=1@localhost:9093 |
There was a problem hiding this comment.
Should this be 3000@localhost:9093?
|
ping @mumrah |
| controller.quorum.voters=3000@localhost:9093 | ||
| controller.listener.names=CONTROLLER</pre> | ||
|
|
||
| <p> |
There was a problem hiding this comment.
Should we mention the errors that are currently logged due to https://issues.apache.org/jira/browse/KAFKA-14698 ?
Something like:
| <p> | |
| <p>During this step both brokers and controllers will start logging errors and warnings. This is an <a href="https://issues.apache.org/jira/browse/KAFKA-14698">issue</a> with the early access migration. | |
| <p>The error on KRaft broker is: | |
| <pre> | |
| org.apache.kafka.common.errors.InvalidRequestException: Received request api key LEADER_AND_ISR which is not enabled | |
| ERROR Closing socket for 192.168.1.11:9092-192.168.1.11:63737-371 because of error (kafka.network.Processor)</pre> | |
| <p>The messages on controllers are: | |
| <pre> | |
| WARN [Controller id=1000, targetBrokerId=0] Connection to node 0 (mmaison-mac.home/192.168.1.11:9092) could not be established. Broker may not be available. (org.apache.kafka.clients.NetworkClient) | |
| INFO [Controller id=1000, targetBrokerId=0] Client requested connection close from node 0 (org.apache.kafka.clients.NetworkClient) | |
| INFO [Controller id=1000, targetBrokerId=0] Node 0 disconnected. (org.apache.kafka.clients.NetworkClient)</pre> | |
| <p>If only these errors appear, you can ignore them and carry on this the migration steps as they don't impact the consistency or health of the cluster. Once the following step is completed, the error should stop. | |
| <p> |
|
@mumrah Let me know if you don't have time to finish this PR. We should merge this to trunk and backport to 3.5. |
|
@mimaison i'm fairly occupied with #13407 and #13461 at the moment, but could probably find time for this PR in between things this week or next. Come to think of it, I'll need to go through the docs before 3.5 to make some adjustments based on recent work (e.g., we now support ACL and soon will support SCRAM migration). |
|
@mimaison can you take another look at this? Sorry for the excessive delay on this one, I'll open a kafka-site PR after this is merged to trunk and 3.5 |
mimaison
left a comment
mimaison
left a comment
There was a problem hiding this comment.
Thanks for the updates. I left a couple of comments
|
|
||
| <p> | ||
| <b>The ZooKeeper to KRaft migration feature is considered Early Access in 3.4.0. It is not recommended for production clusters.</b> | ||
| <b>ZooKeeper to KRaft migration is considered an Early Access feature in 3.4.0 and is not recommended for production clusters.</b> |
There was a problem hiding this comment.
Should we remove in 3.4.0 from this sentence?
| <li>Kafka server's <code>process.role</code> should be set to either <code>broker</code> or <code>controller</code> but not both. Combined mode can be used in development enviroment but it should be avoided in critical deployment evironments.</li> | ||
| <li>For redundancy, a Kafka cluster should use 3 controllers. More than 3 servers is not recommended in critical environments. In the rare case of a partial network failure it is possible for the cluster metadata quorum to become unavailable. This limitation will be addresses in a future release of Kafka.</li> | ||
| <li>Kafka server's <code>process.role</code> should be set to either <code>broker</code> or <code>controller</code> but not both. Combined mode can be used in development environments, but it should be avoided in critical deployment environments.</li> | ||
| <li>For redundancy, a Kafka cluster should use 3 controllers. More than 3 servers is not recommended in critical environments. In the rare case of a partial network failure it is possible for the cluster metadata quorum to become unavailable. This limitation will be addressed in a future release of Kafka.</li> |
There was a problem hiding this comment.
Should More than 3 servers is not recommended in critical environments. be More than 3 controllers is not recommended in critical environments.? Otherwise it could be interpreted as "you can only have 3 brokers"
| <p>The following features are not fully implemented in KRaft mode:</p> | ||
|
|
||
| <ul> | ||
| <li>Configuring SCRAM users via the administrative API</li> |
There was a problem hiding this comment.
Also, can this be removed now?
|
Thanks @mimaison, updated with your suggestions. |
…13257) Reviewers: Mickael Maison <mickael.maison@gmail.com>
2 participants
This patch brings in the ZK migration docs that were added for the 3.4 release.