Fixed
Status Update
No update yet.
Comments
ch...@gmail.com
+1 on this...
I currently have a Cloud Run service running out of memory even at the max of 2GB. Running the same request locally with a memory profiler shows that it consumes ~400MB of memory, so I really have no idea where the discrepancy is coming from.
I currently have a Cloud Run service running out of memory even at the max of 2GB. Running the same request locally with a memory profiler shows that it consumes ~400MB of memory, so I really have no idea where the discrepancy is coming from.
ni...@gmail.com
Thank you for putting together a memory utilization metric. I'm sure this will come in handy.
I have a similar issue right now, trying to run a ML model in app engine. I suspected it was due to local temporary disk usage, since i've seen somewhere that this can count towards your ram budget. I've run my container locally and checked what files changed using "docker diff [container_id]", and the files changed don't account for the discrepancy.
I would be nice to have something closer to a per-process breakdown of memory usage, as well as any other system that could contribute to RAM usage (like ram disk size).
I have a similar issue right now, trying to run a ML model in app engine. I suspected it was due to local temporary disk usage, since i've seen somewhere that this can count towards your ram budget. I've run my container locally and checked what files changed using "docker diff [container_id]", and the files changed don't account for the discrepancy.
I would be nice to have something closer to a per-process breakdown of memory usage, as well as any other system that could contribute to RAM usage (like ram disk size).
gg...@farsightgen.com
+1 on this
The solution offered above is an aggregate measure for large deployment monitoring, for billing and ops management. What is still needed (and the original request) is a within-container measurement tool that can report out during code development and debugging.
The solution offered above is an aggregate measure for large deployment monitoring, for billing and ops management. What is still needed (and the original request) is a within-container measurement tool that can report out during code development and debugging.
qu...@gmail.com
+ 1
sc...@gmail.com
+1
ii...@fund.ar
+1
mi...@gmail.com
+1
ya...@a2xaccounting.com
+1
ta...@iwell.eu
+1
Description
What you would like to accomplish: To be able to monitor the memory usage of Google Cloud Run istances.
How this might work: The metrics for memory usage could be displayed in Stackdriver and on "Cloud Run" page in Google Cloud Console.
If applicable, reasons why alternative solutions are not sufficient: Currently there is no possibility to see the memory usage on "Cloud Run" page in the Google Cloud Console or in Stackdriver - only allocated memory is displayed.
Other information (workarounds you have tried, documentation consulted, etc): N/A