MySQL Forums
Forum List  »  NDB clusters

Durability of transactions in Cluster
Posted by: Sandeep Jangra
Date: October 23, 2015 11:16AM


It seems like when a transaction is committed on cluster it is not written to the disk until one of the checkpoints (LCP or GCP) kicks in. So theoretically there is a very window of time where if all nodes within the node group go down, we can have a potential data loss.

Is there a way to force every transaction to get written to bin logs (not buffers) on commit so that we can avoid the data loss scenario described above?

If the answer to above is YES, then are there any benchmark performance numbers available with those settings?

Please let me know.


Options: ReplyQuote

Written By
Durability of transactions in Cluster
October 23, 2015 11:16AM
November 06, 2015 12:52AM

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.