F1 instances now have 256MB memory?

415 views
Skip to first unread message

Jeff O'Neill

unread,
Dec 31, 2018, 8:51:10 AM12/31/18
to Google App Engine
It seems that F1 instances (at least on GAE standard) now have 256MB memory although the docs still say they have 128MB (https://cloud.google.com/appengine/docs/standard/#instance_classes).

Are there any other changes to the size and speed of instance classes?  I'm thinking about upgrading and want to make sure I have accurate info.

thanks!
Jeff

George (Cloud Platform Support)

unread,
Dec 31, 2018, 3:00:06 PM12/31/18
to Google App Engine
Hello Jeff, 

You might have discovered a discrepancy in different documentation pages. Where did you see 256MB value being indicated? Not aware of any similar changes in F1 values right now. 

What would you like to upgrade to? What specific advantages are you looking for, with this prospective upgrade? 

Jeff O'Neill

unread,
Dec 31, 2018, 4:25:26 PM12/31/18
to google-a...@googlegroups.com
Hi George,

My interest in upgrading is faster processors.  I have some fairly computational tasks that I would like to get done more quickly.

My F1 instances have 256MB.  They regularly go above 128MB and when they exceed 256MB, I get this message in my logs:

Exceeded soft memory limit of 256 MB with 264 MB after servicing...

So despite that the docs say that F1 instances have 128MB, mine actually have 256MB.

Jeff


--
You received this message because you are subscribed to a topic in the Google Groups "Google App Engine" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/google-appengine/OG1Vsif3D5E/unsubscribe.
To unsubscribe from this group and all its topics, send an email to google-appengi...@googlegroups.com.
To post to this group, send email to google-a...@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/5367b2d7-7086-476d-aef0-767b3390c53d%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Tapir

unread,
Jan 1, 2019, 1:08:43 AM1/1/19
to Google App Engine
128M is not enough to run a simple Java web app.

Tiago (Google Cloud Platform Support)

unread,
Jan 4, 2019, 6:45:48 PM1/4/19
to Google App Engine
Hi,

If you’re using the Java runtime on GAE Standard, since its updated from 32 to 64-bit in the Version 1.9.40, the Developer Console indeed reports a higher memory [1] usage for all associated instances. Specifically, small instances show double their sizes (in line with Jeff’s observation above), while larger instances show a smaller increase factor.

Keep in mind that this increase is in off-heap memory. The amount of heap memory that your application can access in F1 or other instances is still in line with what’s described in the official documentation.

[1] https://cloud.google.com/appengine/docs/standard/java/release-notes#july_21_2016

Jeff O'Neill

unread,
Jan 4, 2019, 7:08:30 PM1/4/19
to google-a...@googlegroups.com
Thanks for the explanation.  I'm using Python 2.7 on GAE standard so the same issue must apply there as well.


Tiago (Google Cloud Platform Support)

unread,
Jan 7, 2019, 8:20:10 PM1/7/19
to Google App Engine
Thanks for the clarification as well, Jeff.

The information I provided above only applies to apps deployed using the Java 8 runtime on GAE Standard, though. I’m currently not aware of circumstances when a F1 instance would allocate 256 MB when in use with Python 2.7 on GAE Standard. 

In any case, replying to your initial question, there currently aren't any changes to the size and speed of instance classes on GAE Standard. Apart from the small aside on the Java runtime I explained above, the official documentation [1] indeed provides accurate info regarding how much heap memory and CPU your application has available for usage.

Tiago (Google Cloud Platform Support)

unread,
Jan 8, 2019, 8:00:49 PM1/8/19
to Google App Engine
Let me also add that if what you’re observing regarding the F1 instance memory allocation doesn't correspond to the official documentation provided above, this is usually a sign of a public issue and should be further investigated accordingly. You can open a new issue in the issue tracker here: https://cloud.google.com/support/docs/issue-trackers .
Reply all
Reply to author
Forward
0 new messages