Description:
Currently DUMP 7021 is not documented. DUMP 7021 is the same as DUMP 7015 but for a single table. When debugging specific tables, it can be useful to reduce the output of 7015 to just that/those table(s) in question.
DUMP 7021 takes a single table id as an argument:
shell$ ndb_show_tables
...
id type state logging database schema name
...
12 UserTable Online Yes world def City
shell$ ndb_mgm -e "ALL DUMP 7021 12"
Connected to Management Server at: localhost:1186
Sending dump signal with data:
0x00001b6d 0x0000000c
Sending dump signal with data:
0x00001b6d 0x0000000c
The Cluster log now contains something like:
2016-07-05 09:59:22 [MgmtSrvr] INFO -- Node 1: Table 12: TabCopyStatus: 0 TabUpdateStatus: 0 TabLcpStatus: 3
2016-07-05 09:59:22 [MgmtSrvr] INFO -- Node 1: Fragment 0: noLcpReplicas==0 0(on 1)=32(Idle) 1(on 2)=32(Idle)
2016-07-05 09:59:22 [MgmtSrvr] INFO -- Node 1: Fragment 1: noLcpReplicas==0 0(on 1)=32(Idle) 1(on 2)=32(Idle)
2016-07-05 09:59:23 [MgmtSrvr] INFO -- Node 2: Table 12: TabCopyStatus: 0 TabUpdateStatus: 0 TabLcpStatus: 3
2016-07-05 09:59:23 [MgmtSrvr] INFO -- Node 2: Fragment 0: noLcpReplicas==0 0(on 1)=32(Idle) 1(on 2)=32(Idle)
2016-07-05 09:59:23 [MgmtSrvr] INFO -- Node 2: Fragment 1: noLcpReplicas==0 0(on 1)=32(Idle) 1(on 2)=32(Idle)
How to repeat:
See http://dev.mysql.com/doc/ndbapi/en/ndb-internals-dump-commands.html - DUMP 7021 is missing.
Suggested fix:
Add a page for 7021 and probably change the page for 7015 to note it's the same as 7021 for each table and refer to 7021 for details.
