| Bug #28720 | Disk data meta information is not visible in mysqld but exists in ndbd | ||
|---|---|---|---|
| Submitted: | 28 May 2007 9:35 | Modified: | 11 Jul 2007 9:28 |
| Reporter: | Johan Andersson | Email Updates: | |
| Status: | Closed | Impact on me: | |
| Category: | MySQL Cluster: Disk Data | Severity: | S3 (Non-critical) |
| Version: | mysql-5.1.18 ndb-6.2.3, 5.1.20-BK | OS: | Any |
| Assigned to: | Jonas Oreland | CPU Architecture: | Any |
[30 May 2007 3:19]
Adam Dixon
I have verified this behavior on latest BK of 5.1 - someone with more internals will need to look further as to why. The relevant warning present in Johan's output is as follows; +-------+------+----------------------------------------------------+ | Level | Code | Message | +-------+------+----------------------------------------------------+ | Error | 1296 | Got error 4007 'Send to ndbd node failed' from NDB | +-------+------+----------------------------------------------------+ 1 row in set (0.00 sec) NDB error code 4007: Send to ndbd node failed: Unknown result: Unknown result error As soon as you bring up the second node you get; 9 rows in set (0.00 sec) as expected. IS looks to be bombing out when trying to query/process down node. Where it should continue to display only running nodes files. Repeat process; Start 51 cluster; CREATE LOGFILE GROUP lg_1 ADD UNDOFILE 'undo_1.dat' INITIAL_SIZE 16M UNDO_BUFFER_SIZE 2M ENGINE NDB; ALTER LOGFILE GROUP lg_1 ADD UNDOFILE 'undo_2.dat' INITIAL_SIZE 12M ENGINE NDB; CREATE TABLESPACE ts_1 ADD DATAFILE 'data_1.dat' USE LOGFILE GROUP lg_1 INITIAL_SIZE 32M ENGINE NDB; ALTER TABLESPACE ts_1 ADD DATAFILE 'data_2.dat' INITIAL_SIZE 48M ENGINE NDB; ndb_mgm -e 'all stop' ndbd (wait for partial start) mysql> select * from information_schema.FILES
[4 Jul 2007 10:00]
Jon Stephens
Documented bugfix for telco-6.2.4 release, left status as PQ.
[10 Jul 2007 13:27]
Bugs System
Pushed into 5.1.21-beta
[11 Jul 2007 9:28]
Jon Stephens
Thank you for your bug report. This issue has been committed to our source repository of that product and will be incorporated into the next release.
If necessary, you can access the source repository and build the latest available version, including the bug fix. More information about accessing the source trees is available at
http://dev.mysql.com/doc/en/installing-source.html
Documented bugfix in 5.1.21 changelog.
[6 Sep 2007 9:41]
Jon Stephens
Fix also documented for mysql-5.1.15-ndb-6.1.18.
Description: Disk data files are properly visible from ndb_show_tables, so they seem to exist in cluster: 7 Tablespace Online - ts_1 8 Datafile Online - data_1.dat 5 LogfileGroup Online - lg_1 82 Datafile Online - data_2.dat 6 Undofile Online - undo_1.dat but from mysqld: mysql> select * from information_schema.files\G *************************** 1. row *************************** FILE_ID: 0 FILE_NAME: data_1.dat FILE_TYPE: DATAFILE TABLESPACE_NAME: ts_1 TABLE_CATALOG: NULL TABLE_SCHEMA: NULL TABLE_NAME: NULL LOGFILE_GROUP_NAME: lg_1 LOGFILE_GROUP_NUMBER: NULL ENGINE: ndbcluster FULLTEXT_KEYS: NULL DELETED_ROWS: NULL UPDATE_COUNT: NULL FREE_EXTENTS: 1623 TOTAL_EXTENTS: 12288 EXTENT_SIZE: 1048576 INITIAL_SIZE: 12884901888 MAXIMUM_SIZE: 12884901888 AUTOEXTEND_SIZE: NULL CREATION_TIME: NULL LAST_UPDATE_TIME: NULL LAST_ACCESS_TIME: NULL RECOVER_TIME: NULL TRANSACTION_COUNTER: NULL VERSION: 1 ROW_FORMAT: FIXED TABLE_ROWS: NULL AVG_ROW_LENGTH: NULL DATA_LENGTH: NULL MAX_DATA_LENGTH: NULL INDEX_LENGTH: NULL DATA_FREE: NULL CREATE_TIME: NULL UPDATE_TIME: NULL CHECK_TIME: NULL CHECKSUM: NULL STATUS: NORMAL EXTRA: CLUSTER_NODE=2 1 row in set, 1 warning (0.00 sec) I can only see one data file... How to repeat: No idea yet... :( Found this after doing a partial start of the cluster using version mysql-5.1.18 ndb-6.2.3. Suggested fix: -