Closed ankushshr1993 closed 2 years ago
containerPool.userMemory: total memory available per invoker instance. Invoker uses this memory to create containers for user-actions. The concurrency-limit (actions running in parallel) will depend upon the total memory configured for containerPool and memory allocated per action (default: 256mb per container).
Thanks, I'll check it out.
On Wed, May 4, 2022 at 3:15 AM David Grove @.***> wrote:
containerPool.userMemory: total memory available per invoker instance. Invoker uses this memory to create containers for user-actions. The concurrency-limit (actions running in parallel) will depend upon the total memory configured for containerPool and memory allocated per action (default: 256mb per container).
— Reply to this email directly, view it on GitHub https://github.com/apache/openwhisk-deploy-kube/issues/728#issuecomment-1116847483, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJHA3KVC57MCEGMM3Z6OHJ3VIHFSVANCNFSM5U5N7I3A . You are receiving this because you authored the thread.Message ID: @.***>
Hi David,
As per your suggestion and described in the link I updated the mycluster.yaml file by adding actionsInvokesConcurrent: 50 but it still creates only 8 pods .I am trying to figure out what would be the value of contaierPool.userMemory: forexample for 50 concurrent pods.
I have attached mycluster.yaml file used in my deployment for your reference.I would request for your advice on this.
Regards Ankush Sharma
On Wed, May 4, 2022 at 3:15 AM David Grove @.***> wrote:
containerPool.userMemory: total memory available per invoker instance. Invoker uses this memory to create containers for user-actions. The concurrency-limit (actions running in parallel) will depend upon the total memory configured for containerPool and memory allocated per action (default: 256mb per container).
— Reply to this email directly, view it on GitHub https://github.com/apache/openwhisk-deploy-kube/issues/728#issuecomment-1116847483, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJHA3KVC57MCEGMM3Z6OHJ3VIHFSVANCNFSM5U5N7I3A . You are receiving this because you authored the thread.Message ID: @.***>
You need to increase containerPool.userMemory
. The default value is 2048m, which corresponds to 8 pods of 256m each (the default memory size for a user container). If you want 50 concurrent actions, you have one invoker, and your actions use the default of 256m, then you need containerPool.userMemory = 50*256.
Hi David,
I have tried these values but it stil not working.Can you please have a look at mycluster.yaml file and help if I am keeping it correctly
Regards Ankush Sharma
On Fri, May 6, 2022 at 12:00 AM David Grove @.***> wrote:
You need to increase containerPool.userMemory. The default value is 2048m, which corresponds to 8 pods of 256m each (the default memory size for a user container). If you want 50 concurrent actions, you have one invoker, and your actions use the default of 256m, then you need containerPool.userMemory = 50*256.
— Reply to this email directly, view it on GitHub https://github.com/apache/openwhisk-deploy-kube/issues/728#issuecomment-1119084232, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJHA3KWEBHU2WECPGM56GD3VIRAHZANCNFSM5U5N7I3A . You are receiving this because you authored the thread.Message ID: @.***>
Thanks a lot david for your help
I was putting the variables and values in the wrong section.I figured it out.
Thanks again Regards Ankush Sharma
On Fri, May 6, 2022 at 12:54 AM ANKUSH SHARMA @.***> wrote:
Hi David,
I have tried these values but it stil not working.Can you please have a look at mycluster.yaml file and help if I am keeping it correctly
Regards Ankush Sharma
On Fri, May 6, 2022 at 12:00 AM David Grove @.***> wrote:
You need to increase containerPool.userMemory. The default value is 2048m, which corresponds to 8 pods of 256m each (the default memory size for a user container). If you want 50 concurrent actions, you have one invoker, and your actions use the default of 256m, then you need containerPool.userMemory = 50*256.
— Reply to this email directly, view it on GitHub https://github.com/apache/openwhisk-deploy-kube/issues/728#issuecomment-1119084232, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJHA3KWEBHU2WECPGM56GD3VIRAHZANCNFSM5U5N7I3A . You are receiving this because you authored the thread.Message ID: @.***>
Openwhisk allows only 8 pods to be created for n number of invocations at a time. For performance testing, we need to increase the number of pods at a time.Is this a limitation of open whisk or can configure it to eradicate this limitation