Instance scalability - high volume of events and thousands of users

[ceskie]

Ahoy,

I'd like to ask/search for someone who managed to has the instance's performance (available memory, number of processing nodes, etc.) upscaled. Our instance is suffering from low resources. The instance's performance is - judging by the performance dashboard - not meeting our expectations approx 85% of the day. 

Some numbers:
we have up to ~3k users active daily
we process 6 - 10 M of events a day
we are hosting 116 scoped apps and utilizing a handful of global plugins (ITSM, ITOM, HR, ...)

Some examples of delay:
During peak hours, incoming email event takes 2-3 hours to process (this is crazy, customer asks us for action and we receive the email with 3h delay and then we act???)

Wording from support: instance is not in trouble as the events are being processed
- well yes, they are, but hours after they are created

Any idea how to force ServiceNow to grant us more Hardware? I am afraid otherwise we will have to declare ServiceNow is not fit for purpose and we'd have to move some of out solutions outside the platform.

We have already involved Now Support multiple times and performed multiple refactoring (costing us 100ks of $) with little to no effect.

Thank you

[Jeff Currier]

Hardware adjustments should be part of the performance case so they should be requested or recommended there. Not sure why your previous cases costed you $$s

[ceskie]

Ahoy, thank you, the cases itself did not cost us $, the (multiple rounds of) refactoring did. Can you please ellaborate a little the recommendation of HW upgrade based on the performance case? Do you have any experience with such a case from the past please?

[Hitoshi Ozawa]

Hi,

If the problem is just with email, can it be the smtp server. Check the smtp queue length and time in queue.

[ceskie]

Ahoy, thank you but unfortunately its problem with all events not just email processing