Bug #49807 2nd management node refuses to start when 1st mgmd and data nodes are running
Submitted: 18 Dec 2009 15:22 Modified: 28 Dec 2009 20:21
Reporter: Victoria Reznichenko Email Updates:
Status: Closed Impact on me:
None 
Category:MySQL Cluster: Cluster (NDB) storage engine Severity:S3 (Non-critical)
Version:7.0.10 OS:Any
Assigned to: Magnus Blåudd CPU Architecture:Any

[18 Dec 2009 15:22] Victoria Reznichenko 
Description:
If one management node and data nodes are running and 2nd management node fetches config from 1st management node, you get the following error on 2nd management node:

-bash-3.2$ ./libexec/ndb_mgmd -c 127.0.0.1:1190 --ndb-nodeid=2 --initial --nodaemon --verbose
2009-12-18 17:11:08 [MgmtSrvr] INFO     -- NDB Cluster Management Server. mysql-5.1.39 ndb-7.0.10
2009-12-18 17:11:08 [MgmtSrvr] DEBUG    -- Got nodeid: 2 from command line or NDB_CONNECTSTRING
2009-12-18 17:11:08 [MgmtSrvr] DEBUG    -- Deleting binary config file '/data2/users/victoria/mysql-cluster-7.0.10/mysql-cluster/ndb_2_config.bin.1'
2009-12-18 17:11:08 [MgmtSrvr] INFO     -- Trying to get configuration from other mgmd(s) using 'nodeid=2,127.0.0.1:1190'...
2009-12-18 17:11:08 [MgmtSrvr] INFO     -- Connected to '127.0.0.1:1190'...
2009-12-18 17:11:08 [MgmtSrvr] ERROR    -- INTERNAL ERROR: Found dynamic ports with value in config, n1: 3, n2: 4, port: 4294909670
2009-12-18 17:11:08 [MgmtSrvr] ERROR    -- INTERNAL ERROR: Found dynamic ports with value in config, n1: 17, n2: 3, port: 4294922349
2009-12-18 17:11:08 [MgmtSrvr] ERROR    -- INTERNAL ERROR: Found dynamic ports with value in config, n1: 17, n2: 4, port: 4294921379
2009-12-18 17:11:08 [MgmtSrvr] ERROR    -- INTERNAL ERROR: Found dynamic ports with value in config, n1: 18, n2: 3, port: 4294930700
2009-12-18 17:11:08 [MgmtSrvr] ERROR    -- INTERNAL ERROR: Found dynamic ports with value in config, n1: 18, n2: 4, port: 4294912796
ConfigManager.cpp:609: require((ok)) failed
Aborted

How to repeat:
1. set up test cluster with 2 management nodes
2. run 1st management node with --nowait-nodes option
3. start data nodes
4. start 2nd management node so it fetches configuration file from 1st management node
[21 Dec 2009 11:00] Bugs System 
A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/95203
[22 Dec 2009 17:53] Magnus Blåudd 
Pushed to 7.0
[28 Dec 2009 20:21] Jon Stephens 
Documented bugfix in the NDB-7.0.11 changelog as follows:

        When attempting to join a running cluster whose management
        server had been started with --nowait-nodes and having SQL nodes
        with dynamically allocated node IDs, a second management server
        failed with spurious INTERNAL ERROR: Found dynamic ports with
        value in config... errors.

Closed.