MySQL Forums
Forum List  »  NDB clusters

Re: Error with "Read Backup"
Posted by: Jeff Steele
Date: August 31, 2017 10:52AM

Reorganizing the partitions first worked, with one caveat. The query was actually:

ALTER TABLE < Table Name > algorithm=inplace, REORGANIZE PARTITION

with a singular "PARTITION".

However, after reorganizing the partitions and setting "NDB_TABLE=READ_BACKUP=1", the ndbmtd started using lots of resources, driving CPU usage to over 500%. This causes the sql nodes to work so slowly that my web application timed out and was unusable. I then set "NDB_TABLE=READ_BACKUP=0" to see if things would go back to normal, but so far they haven't.

You noticed a discrepancy between my NoOfFragmentLogParts setting and MaxNoOfExecutionThreads. Should I change that in some way?

Options: ReplyQuote


Subject
Views
Written By
Posted
1031
August 29, 2017 03:25PM
549
August 29, 2017 03:50PM
469
August 30, 2017 08:52AM
561
August 31, 2017 03:15AM
503
August 31, 2017 04:30AM
Re: Error with "Read Backup"
586
August 31, 2017 10:52AM
549
August 31, 2017 12:41PM
617
August 31, 2017 01:57PM
628
August 31, 2017 02:00PM
587
August 31, 2017 04:37PM
570
August 31, 2017 04:59PM


Sorry, you can't reply to this topic. It has been closed.

Content reproduced on this site is the property of the respective copyright holders. It is not reviewed in advance by Oracle and does not necessarily represent the opinion of Oracle or any other party.