Scaling writes google cloud sql

[Susan Lin]

Is there any guides how to scale writes in Google SQL? It would be great if there was a step by step tutorial how to scale writes using Google SQL.

[Brian Wawok]

Writes to your instance should be proportional to how much CPU and DISK IO you have, right?  So the bigger your instance and the bigger your storage, the more CPU and DISK IO you have.

Generally the more indexes you have, and the more complicated your Primary Key is, the slower your write will be..  A single autoinc primary key and no other indexes or FKs would maximize your performance...

Brian

[Susan Lin]

Thanks for the reply. What if the the performance needed exceeds that of the capbilities of a single instance and sharing is needed, is there a guide how go set this up on Google SQL.

[Brian Wawok]

Sharding?  How many writes are you looking at? How big of writes?

Etsy has a bunch of articles on Sharding MySQL.. for example:

http://www.slideshare.net/jgoulah/the-etsy-shard-architecture-starts-with-s-and-ends-with-hard

You mostly deal with it in your app logic, so nothing to do with Cloud SQL.

Or if you don't want to do that, you may be too big for MySQL, but Etsy is pretty big...

Brian

On Sat, Dec 26, 2015 at 2:13 PM Susan Lin <susan...@gmail.com> wrote:

Thanks for the reply. What if the the performance needed exceeds that of the capbilities of a single instance and sharing is needed, is there a guide how go set this up on Google SQL.

--
You received this message because you are subscribed to a topic in the Google Groups "Google Cloud SQL discuss" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/google-cloud-sql-discuss/wmjT35IOikk/unsubscribe.
To unsubscribe from this group and all its topics, 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/e9c482c8-108e-42c4-94cf-eadbb0d1c93c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Chris Lewis]

We would generally recommend you create read replicas, and then place as much read-only load that you can onto the replicas. That way you free up resources on your master (which is now only handling writes) and should now be able to handle more write traffic.

If you can live without an SLA and have the opportunity to migrate over, a second generation instance will be much faster at writes than a first generation instance will be.

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/CAJVyzE1yvJy7S68Rgp3eX0rZ2DH7Ccm69%3DMfd5Hi9uKDw%3DbkZA%40mail.gmail.com.