Hi Google Team,
I have recently migrated an application from old Managed VMs to "flex". My
application does server side processing using flex and we use servers in
24x7 mode. After few days of using flex, I find it way more expensive than
alternatives.
Here are few observations.
- For Same VM configuration: Flex pricing is nearly 3 times for my use case
after including just Compute Engine sustained pricing. Even the
"vm/core/hour" or "RAM/hour" costs are way more.
- Datastore costs have gone up much more because of missing "Memcache" in
Flex. Even with Cloud Redis, we are paying additional cost.
- We had to rewrite entire codebase which was using messaging to PubSub
In effect, we don't see the additional value in Flex that we are being
charged for. In fact, much of the value has been removed. I understand
reduced DevOps, Health Check with flex is valuable, however, the pricing of
it should be reconsidered. At this point of time, I am considering either
to move off entirely from AppEngine or atleast moving to our flex instance
to Google Kubernetes Engine.
Few questions that would help us decide:
- Are there any considerations to make Flex pricing fair in the coming
weeks?
- How reliable is this page giving an example of migration of Flex ->
GKE
:
https://cloud.google.com/appengine/docs/flexible/python/run-flex-app-on-kubernetes
. Are there any examples in Github?
- What is your Roadmap regarding CloudTasks, Memcache for Flex?
Regards,
Rahul Tongia
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to google-appengine+unsubscr...@googlegroups.com.
To post to this group, send email to google-appengine@googlegroups.com.
Visit this group at https://groups.google.com/group/google-appengine.
To view this discussion on the web visit
https://groups.google.com/d/msgid/google-appengine/d9b06595-e0ca-4975-8187-2fb04018fc2f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.