MySQL Second Generation read replica downtime?

[徐弘宇]

 I used MySQL Second Generation read replica, as slave DB, but it suddenly going maintained .

I have only one read replica, it impacts my services, how to I avoid this situation?

Any solutions or practice will be helpful, thanks.

[Aref Amiri (Cloud Platform Support)]

This could happen if you delete a read replica and create a new one with the same name. If that is not the case, I would suggest to check QPS ratio on your master instance, as it could overload the read replica, which in this case a good practice would be resharding your master instance to 2 or more instances to split the query load. You can refer to more best practices here.

To further assist you with the issue, I would suggest to create a private issue on Public Issue Tracker page and we will be able to further investigate it with our internal tools and provide you with more meaningful information.

[Nick Pomfret]

Thanks for the response.  There are no read replicas here, it's just a single master instance.  I'll open a ticket.

On Thu, 5 Sep 2019 at 18:34, 'Aref Amiri (Cloud Platform Support)' via Google Cloud SQL discuss <google-cloud...@googlegroups.com> wrote:

This could happen if you delete a read replica and create a new one with the same name. If that is not the case, I would suggest to check QPS ratio on your master instance, as it could overload the read replica, which in this case a good practice would be resharding your master instance to 2 or more instances to split the query load. You can refer to more best practices here.

To further assist you with the issue, I would suggest to create a private issue on Public Issue Tracker page and we will be able to further investigate it with our internal tools and provide you with more meaningful information.

--
You received this message because you are subscribed to the Google Groups "Google Cloud SQL discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to google-cloud-sql-d...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/google-cloud-sql-discuss/550d651e-8470-4411-b558-a3ed2e307cbf%40googlegroups.com.

[Aref Amiri (Cloud Platform Support)]

Hi Nick,

The suggestion posted here is related to the issue initially posted about a read replica instance going into maintenance mode. 

I have also posted some suggestions pertaining the latency issue you are experiencing on you Cloud SQL instance here. Please refer to the other thread and let us know if you have other questions.