we are planning to use AEM6.2 for communities. Could you please provide me your thoughts for best AEM6.2 capacity and Volume(CPUs, Ram, Heapsize, etc..) or Infrastructure needed for 50-100 concurrent users. I know we have to consider caching requirements, designs, etc(Assume average architecture provided).
thanks
raj
Solved! Go to Solution.
Eng replied with this information:
For communities Authoring environment – Sizing guidelines are same as Sites. [0]
Being Write heavy and UGC heavy, publish environment should follow the recommended cloud service architecture for communities by either using ASRP ( cloud storage ) OR MSRP/DSRP ( on prem installations). Using JSRP is not a recommended deployment for production setup.
Publish Farm is a better alternative with user sync configured. UGC would be automatically sync across instances in case of SRPs ( except JSRP).
For 50-100 concurrent users .. not sure if these are actual concurrent UGC write requests or just read. But heap size of more than 16GB is least we should consider for a publish instance.
For more details on topologies refer [1]
For publish instance hardware sizing calculation , refer to [2] and use caching % almost 10 % or almost 0 for communtites use case.
[0] https://helpx.adobe.com/experience-manager/6-2/managing/using/hardware-sizing-guidelines.html
[1] https://helpx.adobe.com/experience-manager/6-2/communities/using/topologies.html
[2] https://helpx.adobe.com/experience-manager/6-2/managing/using/hardware-sizing-guidelines.html
I have asked our team to look at this question.
Views
Replies
Total Likes
Thank you so much for the quick response.
thanks
raj
Views
Replies
Total Likes
Eng replied with this information:
For communities Authoring environment – Sizing guidelines are same as Sites. [0]
Being Write heavy and UGC heavy, publish environment should follow the recommended cloud service architecture for communities by either using ASRP ( cloud storage ) OR MSRP/DSRP ( on prem installations). Using JSRP is not a recommended deployment for production setup.
Publish Farm is a better alternative with user sync configured. UGC would be automatically sync across instances in case of SRPs ( except JSRP).
For 50-100 concurrent users .. not sure if these are actual concurrent UGC write requests or just read. But heap size of more than 16GB is least we should consider for a publish instance.
For more details on topologies refer [1]
For publish instance hardware sizing calculation , refer to [2] and use caching % almost 10 % or almost 0 for communtites use case.
[0] https://helpx.adobe.com/experience-manager/6-2/managing/using/hardware-sizing-guidelines.html
[1] https://helpx.adobe.com/experience-manager/6-2/communities/using/topologies.html
[2] https://helpx.adobe.com/experience-manager/6-2/managing/using/hardware-sizing-guidelines.html
Views
Likes
Replies
Views
Like
Replies