You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by zentol <gi...@git.apache.org> on 2018/02/12 14:28:16 UTC

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

GitHub user zentol opened a pull request:

    https://github.com/apache/flink/pull/5462

    [FLINK-8475][config][docs] Integrate HA-ZK options

    ## What is the purpose of the change
    
    This PR integrates the Zookeeper HA `ConfigOptions` into the configuration docs generator.
    
    ## Brief change log
    
    * Add `ConfigGroup` for zookeeper HA config options
    * integrate zookeeper HA configuration table into `config.md`

You can merge this pull request into a Git repository by running:

    $ git pull https://github.com/zentol/flink 8475_ha_zk

Alternatively you can review and apply these changes as the patch at:

    https://github.com/apache/flink/pull/5462.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

    This closes #5462
    
----
commit 14e168c1c9c526a5feb0cac5738ea9a3663b0466
Author: zentol <ch...@...>
Date:   2018-01-23T12:50:32Z

    [FLINK-8475][config][docs] Integrate HA-ZK options

----


---

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by aljoscha <gi...@git.apache.org>.
Github user aljoscha commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5462#discussion_r167603434
  
    --- Diff: docs/ops/config.md ---
    @@ -502,27 +502,7 @@ Previously this key was named `recovery.mode` and the default value was `standal
     
     #### ZooKeeper-based HA Mode
     
    -- `high-availability.zookeeper.quorum`: Defines the ZooKeeper quorum URL which is used to connect to the ZooKeeper cluster when the 'zookeeper' HA mode is selected. Previously this key was named `recovery.zookeeper.quorum`.
    -
    -- `high-availability.zookeeper.path.root`: (Default `/flink`) Defines the root dir under which the ZooKeeper HA mode will create namespace directories. Previously this ket was named `recovery.zookeeper.path.root`.
    -
    -- `high-availability.zookeeper.path.latch`: (Default `/leaderlatch`) Defines the znode of the leader latch which is used to elect the leader. Previously this key was named `recovery.zookeeper.path.latch`.
    -
    -- `high-availability.zookeeper.path.leader`: (Default `/leader`) Defines the znode of the leader which contains the URL to the leader and the current leader session ID. Previously this key was named `recovery.zookeeper.path.leader`.
    -
    -- `high-availability.storageDir`: Defines the directory in the state backend where the JobManager metadata will be stored (ZooKeeper only keeps pointers to it). Required for HA. Previously this key was named `recovery.zookeeper.storageDir` and `high-availability.zookeeper.storageDir`.
    --- End diff --
    
    Did you find out why this is gone?


---

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by zentol <gi...@git.apache.org>.
Github user zentol commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5462#discussion_r167610029
  
    --- Diff: docs/ops/config.md ---
    @@ -502,27 +502,7 @@ Previously this key was named `recovery.mode` and the default value was `standal
     
     #### ZooKeeper-based HA Mode
     
    -- `high-availability.zookeeper.quorum`: Defines the ZooKeeper quorum URL which is used to connect to the ZooKeeper cluster when the 'zookeeper' HA mode is selected. Previously this key was named `recovery.zookeeper.quorum`.
    -
    -- `high-availability.zookeeper.path.root`: (Default `/flink`) Defines the root dir under which the ZooKeeper HA mode will create namespace directories. Previously this ket was named `recovery.zookeeper.path.root`.
    -
    -- `high-availability.zookeeper.path.latch`: (Default `/leaderlatch`) Defines the znode of the leader latch which is used to elect the leader. Previously this key was named `recovery.zookeeper.path.latch`.
    -
    -- `high-availability.zookeeper.path.leader`: (Default `/leader`) Defines the znode of the leader which contains the URL to the leader and the current leader session ID. Previously this key was named `recovery.zookeeper.path.leader`.
    -
    -- `high-availability.storageDir`: Defines the directory in the state backend where the JobManager metadata will be stored (ZooKeeper only keeps pointers to it). Required for HA. Previously this key was named `recovery.zookeeper.storageDir` and `high-availability.zookeeper.storageDir`.
    -
    -- `high-availability.zookeeper.client.session-timeout`: (Default `60000`) Defines the session timeout for the ZooKeeper session in ms. Previously this key was named `recovery.zookeeper.client.session-timeout`
    -
    -- `high-availability.zookeeper.client.connection-timeout`: (Default `15000`) Defines the connection timeout for ZooKeeper in ms. Previously this key was named `recovery.zookeeper.client.connection-timeout`.
    -
    -- `high-availability.zookeeper.client.retry-wait`: (Default `5000`) Defines the pause between consecutive retries in ms. Previously this key was named `recovery.zookeeper.client.retry-wait`.
    -
    -- `high-availability.zookeeper.client.max-retry-attempts`: (Default `3`) Defines the number of connection retries before the client gives up. Previously this key was named `recovery.zookeeper.client.max-retry-attempts`.
    -
    -- `high-availability.job.delay`: (Default `akka.ask.timeout`) Defines the delay before persisted jobs are recovered in case of a master recovery situation. Previously this key was named `recovery.job.delay`.
    --- End diff --
    
    ah, this is a general HA option and not specifically tied to ZK. It should be moved into the HA section instead of outright removing it.


---

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by aljoscha <gi...@git.apache.org>.
Github user aljoscha commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5462#discussion_r167603044
  
    --- Diff: docs/ops/config.md ---
    @@ -502,27 +502,7 @@ Previously this key was named `recovery.mode` and the default value was `standal
     
     #### ZooKeeper-based HA Mode
     
    -- `high-availability.zookeeper.quorum`: Defines the ZooKeeper quorum URL which is used to connect to the ZooKeeper cluster when the 'zookeeper' HA mode is selected. Previously this key was named `recovery.zookeeper.quorum`.
    -
    -- `high-availability.zookeeper.path.root`: (Default `/flink`) Defines the root dir under which the ZooKeeper HA mode will create namespace directories. Previously this ket was named `recovery.zookeeper.path.root`.
    --- End diff --
    
    `ket` ... 😅 


---

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by zentol <gi...@git.apache.org>.
Github user zentol commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5462#discussion_r167609071
  
    --- Diff: docs/ops/config.md ---
    @@ -502,27 +502,7 @@ Previously this key was named `recovery.mode` and the default value was `standal
     
     #### ZooKeeper-based HA Mode
     
    -- `high-availability.zookeeper.quorum`: Defines the ZooKeeper quorum URL which is used to connect to the ZooKeeper cluster when the 'zookeeper' HA mode is selected. Previously this key was named `recovery.zookeeper.quorum`.
    -
    -- `high-availability.zookeeper.path.root`: (Default `/flink`) Defines the root dir under which the ZooKeeper HA mode will create namespace directories. Previously this ket was named `recovery.zookeeper.path.root`.
    -
    -- `high-availability.zookeeper.path.latch`: (Default `/leaderlatch`) Defines the znode of the leader latch which is used to elect the leader. Previously this key was named `recovery.zookeeper.path.latch`.
    -
    -- `high-availability.zookeeper.path.leader`: (Default `/leader`) Defines the znode of the leader which contains the URL to the leader and the current leader session ID. Previously this key was named `recovery.zookeeper.path.leader`.
    -
    -- `high-availability.storageDir`: Defines the directory in the state backend where the JobManager metadata will be stored (ZooKeeper only keeps pointers to it). Required for HA. Previously this key was named `recovery.zookeeper.storageDir` and `high-availability.zookeeper.storageDir`.
    -
    -- `high-availability.zookeeper.client.session-timeout`: (Default `60000`) Defines the session timeout for the ZooKeeper session in ms. Previously this key was named `recovery.zookeeper.client.session-timeout`
    -
    -- `high-availability.zookeeper.client.connection-timeout`: (Default `15000`) Defines the connection timeout for ZooKeeper in ms. Previously this key was named `recovery.zookeeper.client.connection-timeout`.
    -
    -- `high-availability.zookeeper.client.retry-wait`: (Default `5000`) Defines the pause between consecutive retries in ms. Previously this key was named `recovery.zookeeper.client.retry-wait`.
    -
    -- `high-availability.zookeeper.client.max-retry-attempts`: (Default `3`) Defines the number of connection retries before the client gives up. Previously this key was named `recovery.zookeeper.client.max-retry-attempts`.
    -
    -- `high-availability.job.delay`: (Default `akka.ask.timeout`) Defines the delay before persisted jobs are recovered in case of a master recovery situation. Previously this key was named `recovery.job.delay`.
    --- End diff --
    
    ehhh....could be that the tables are outdated (ironic isn't it), let me check.


---

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by zentol <gi...@git.apache.org>.
Github user zentol commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5462#discussion_r167793610
  
    --- Diff: docs/ops/config.md ---
    @@ -502,27 +502,7 @@ Previously this key was named `recovery.mode` and the default value was `standal
     
     #### ZooKeeper-based HA Mode
     
    -- `high-availability.zookeeper.quorum`: Defines the ZooKeeper quorum URL which is used to connect to the ZooKeeper cluster when the 'zookeeper' HA mode is selected. Previously this key was named `recovery.zookeeper.quorum`.
    -
    -- `high-availability.zookeeper.path.root`: (Default `/flink`) Defines the root dir under which the ZooKeeper HA mode will create namespace directories. Previously this ket was named `recovery.zookeeper.path.root`.
    -
    -- `high-availability.zookeeper.path.latch`: (Default `/leaderlatch`) Defines the znode of the leader latch which is used to elect the leader. Previously this key was named `recovery.zookeeper.path.latch`.
    -
    -- `high-availability.zookeeper.path.leader`: (Default `/leader`) Defines the znode of the leader which contains the URL to the leader and the current leader session ID. Previously this key was named `recovery.zookeeper.path.leader`.
    -
    -- `high-availability.storageDir`: Defines the directory in the state backend where the JobManager metadata will be stored (ZooKeeper only keeps pointers to it). Required for HA. Previously this key was named `recovery.zookeeper.storageDir` and `high-availability.zookeeper.storageDir`.
    -
    -- `high-availability.zookeeper.client.session-timeout`: (Default `60000`) Defines the session timeout for the ZooKeeper session in ms. Previously this key was named `recovery.zookeeper.client.session-timeout`
    -
    -- `high-availability.zookeeper.client.connection-timeout`: (Default `15000`) Defines the connection timeout for ZooKeeper in ms. Previously this key was named `recovery.zookeeper.client.connection-timeout`.
    -
    -- `high-availability.zookeeper.client.retry-wait`: (Default `5000`) Defines the pause between consecutive retries in ms. Previously this key was named `recovery.zookeeper.client.retry-wait`.
    -
    -- `high-availability.zookeeper.client.max-retry-attempts`: (Default `3`) Defines the number of connection retries before the client gives up. Previously this key was named `recovery.zookeeper.client.max-retry-attempts`.
    -
    -- `high-availability.job.delay`: (Default `akka.ask.timeout`) Defines the delay before persisted jobs are recovered in case of a master recovery situation. Previously this key was named `recovery.job.delay`.
    --- End diff --
    
    re-added in #5467


---

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by aljoscha <gi...@git.apache.org>.
Github user aljoscha commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5462#discussion_r167603849
  
    --- Diff: docs/ops/config.md ---
    @@ -502,27 +502,7 @@ Previously this key was named `recovery.mode` and the default value was `standal
     
     #### ZooKeeper-based HA Mode
     
    -- `high-availability.zookeeper.quorum`: Defines the ZooKeeper quorum URL which is used to connect to the ZooKeeper cluster when the 'zookeeper' HA mode is selected. Previously this key was named `recovery.zookeeper.quorum`.
    -
    -- `high-availability.zookeeper.path.root`: (Default `/flink`) Defines the root dir under which the ZooKeeper HA mode will create namespace directories. Previously this ket was named `recovery.zookeeper.path.root`.
    -
    -- `high-availability.zookeeper.path.latch`: (Default `/leaderlatch`) Defines the znode of the leader latch which is used to elect the leader. Previously this key was named `recovery.zookeeper.path.latch`.
    -
    -- `high-availability.zookeeper.path.leader`: (Default `/leader`) Defines the znode of the leader which contains the URL to the leader and the current leader session ID. Previously this key was named `recovery.zookeeper.path.leader`.
    -
    -- `high-availability.storageDir`: Defines the directory in the state backend where the JobManager metadata will be stored (ZooKeeper only keeps pointers to it). Required for HA. Previously this key was named `recovery.zookeeper.storageDir` and `high-availability.zookeeper.storageDir`.
    -
    -- `high-availability.zookeeper.client.session-timeout`: (Default `60000`) Defines the session timeout for the ZooKeeper session in ms. Previously this key was named `recovery.zookeeper.client.session-timeout`
    -
    -- `high-availability.zookeeper.client.connection-timeout`: (Default `15000`) Defines the connection timeout for ZooKeeper in ms. Previously this key was named `recovery.zookeeper.client.connection-timeout`.
    -
    -- `high-availability.zookeeper.client.retry-wait`: (Default `5000`) Defines the pause between consecutive retries in ms. Previously this key was named `recovery.zookeeper.client.retry-wait`.
    -
    -- `high-availability.zookeeper.client.max-retry-attempts`: (Default `3`) Defines the number of connection retries before the client gives up. Previously this key was named `recovery.zookeeper.client.max-retry-attempts`.
    -
    -- `high-availability.job.delay`: (Default `akka.ask.timeout`) Defines the delay before persisted jobs are recovered in case of a master recovery situation. Previously this key was named `recovery.job.delay`.
    --- End diff --
    
    And also where this is gone?


---

[GitHub] flink issue #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by aljoscha <gi...@git.apache.org>.
Github user aljoscha commented on the issue:

    https://github.com/apache/flink/pull/5462
  
    Ok, looks good now. 👍 


---

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by zentol <gi...@git.apache.org>.
Github user zentol commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5462#discussion_r167608682
  
    --- Diff: docs/ops/config.md ---
    @@ -502,27 +502,7 @@ Previously this key was named `recovery.mode` and the default value was `standal
     
     #### ZooKeeper-based HA Mode
     
    -- `high-availability.zookeeper.quorum`: Defines the ZooKeeper quorum URL which is used to connect to the ZooKeeper cluster when the 'zookeeper' HA mode is selected. Previously this key was named `recovery.zookeeper.quorum`.
    -
    -- `high-availability.zookeeper.path.root`: (Default `/flink`) Defines the root dir under which the ZooKeeper HA mode will create namespace directories. Previously this ket was named `recovery.zookeeper.path.root`.
    -
    -- `high-availability.zookeeper.path.latch`: (Default `/leaderlatch`) Defines the znode of the leader latch which is used to elect the leader. Previously this key was named `recovery.zookeeper.path.latch`.
    -
    -- `high-availability.zookeeper.path.leader`: (Default `/leader`) Defines the znode of the leader which contains the URL to the leader and the current leader session ID. Previously this key was named `recovery.zookeeper.path.leader`.
    -
    -- `high-availability.storageDir`: Defines the directory in the state backend where the JobManager metadata will be stored (ZooKeeper only keeps pointers to it). Required for HA. Previously this key was named `recovery.zookeeper.storageDir` and `high-availability.zookeeper.storageDir`.
    --- End diff --
    
    subsumed by `"high-availability.storageDir"`


---

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by zentol <gi...@git.apache.org>.
Github user zentol commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5462#discussion_r167793417
  
    --- Diff: docs/ops/config.md ---
    @@ -502,27 +502,7 @@ Previously this key was named `recovery.mode` and the default value was `standal
     
     #### ZooKeeper-based HA Mode
     
    -- `high-availability.zookeeper.quorum`: Defines the ZooKeeper quorum URL which is used to connect to the ZooKeeper cluster when the 'zookeeper' HA mode is selected. Previously this key was named `recovery.zookeeper.quorum`.
    -
    -- `high-availability.zookeeper.path.root`: (Default `/flink`) Defines the root dir under which the ZooKeeper HA mode will create namespace directories. Previously this ket was named `recovery.zookeeper.path.root`.
    -
    -- `high-availability.zookeeper.path.latch`: (Default `/leaderlatch`) Defines the znode of the leader latch which is used to elect the leader. Previously this key was named `recovery.zookeeper.path.latch`.
    -
    -- `high-availability.zookeeper.path.leader`: (Default `/leader`) Defines the znode of the leader which contains the URL to the leader and the current leader session ID. Previously this key was named `recovery.zookeeper.path.leader`.
    -
    -- `high-availability.storageDir`: Defines the directory in the state backend where the JobManager metadata will be stored (ZooKeeper only keeps pointers to it). Required for HA. Previously this key was named `recovery.zookeeper.storageDir` and `high-availability.zookeeper.storageDir`.
    --- End diff --
    
    scratch the previous comment (wasn't too helpful was it :D ); this option will now be listed in the general HA section.


---

[GitHub] flink issue #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by zentol <gi...@git.apache.org>.
Github user zentol commented on the issue:

    https://github.com/apache/flink/pull/5462
  
    merging.


---

[GitHub] flink pull request #5462: [FLINK-8475][config][docs] Integrate HA-ZK options

Posted by asfgit <gi...@git.apache.org>.
Github user asfgit closed the pull request at:

    https://github.com/apache/flink/pull/5462


---