How to interpret EwsCutoffBalance in an Exchange throttling policy?

Question

I try to figure out how throttling policies affect EWS.

For EWS, we have these values:

• EWSMaxSubscription: the number of active subscriptions done by the impersonated user.
• EWSMaxConcurrency: how many concurrent connections or actions a single client may take.
• EwsMaxBurst: how far above the standard resource limit a client may go in shorts bursts (in milliseconds). It probably comes into effect when the percentage of CPU/Memory usage by Exchange exceeds the defined threshold (depending of the setup, I suppose).
• EwsRechargeRate: the speed at which the user’s resource budget recharges or refills (in milliseconds).

I understand each of the above throttling parameters. However, I'm not sure to clearly understand the EwsCutoffBalance. This parameter defines the resource consumption limits for EWS user before that user is completely blocked from performing operations on a specific component...

My questions...

1. How this value is used regarding the EwsMaxBurst and the EwsRechargeRate values?
2. What is the unit of this parameter?
3. How can we determine the right value if I need to change the throttling policy of a specific user account (instead of using "Unlimited")?

Answer 1

Both EwsMaxBurst and the EwsRechargeRate are expressed in ms. By default, client will be blocked after 5 minutes (300000 / 1000ms /60s) heavy usage. When recharge rate is 900000 by default, system will recharge in 15 minutes. If you like to keep client requests processed, you can either increase maxburst over recharge rate or decrease recharge rate less than maxburst