Re: Data node deletion in NDB

[Tayyab Fayyaz]

Hi Mauritz Sundell,

Thank you for your input.

I found some user tables and ordered index part if 3,4 nods id, My question here drop the object so other node groups (3,4) will not cause the data loss?


Can I avoid the drop table statement and move all the fragments to other nodes?



SELECT DISTINCT node_id, fq_name, type FROM ndbinfo.memory_per_fragment WHERE node_id IN (3,4);
+---------+-----------------------------+---------------+
| node_id | fq_name | type |
+---------+-----------------------------+---------------+
| 3 | test1/def/test1_retry_queue | User table |
| 4 | test1/def/test1_retry_queue | User table |
| 3 | test1/def/test1 | User table |
| 4 | test1/def/test1 | User table |
| 3 | test1/def/NDB$BLOB_13_13 | User table |
| 4 | test1/def/NDB$BLOB_13_13 | User table |
| 3 | test1/def/test1 | User table |
| 4 | test1/def/test1 | User table |
| 3 | test1/def/NDB$BLOB_16_11 | User table |
| 4 | test1/def/NDB$BLOB_16_11 | User table |
| 3 | sys/def/13/PRIMARY | Ordered index |
| 4 | sys/def/13/PRIMARY | Ordered index |
| 3 | sys/def/16/PRIMARY | Ordered index |
| 4 | sys/def/16/PRIMARY | Ordered index |
+---------+-----------------------------+---------------+