- Application time-outs that are caused by an EWS throttling delay
- Denied connections that are caused by EWS concurrent connection throttling
- "500" or "503" HTTP errors
- A fundamental architecture change in Exchange Server 2013 causes the EWS throttling budget to be calculated based on the back-end (Mailbox) servers instead of on the Client Access server (as it is done in Exchange Server 2010).
- When you use a service account to access various mailboxes, the connections are routed to the back-end server that hosts the target mailbox. However, the default connection-routing path is made to the back-end server that hosts the service account mailbox. Therefore, the EWS throttling budget is calculated based on the back-end server that hosts the service account mailbox, regardless of which back-end server hosts the target mailbox.
Force the Exchange front-end server to route the EWS connections to the back-end server that hosts the target mailbox. To do this, implement the X-AnchorMailbox HTTP header and its value in your EWS calls. The value of the header is defined as the SMTP address of the target mailbox that's accessed.
In the EWS-managed API, use the Add method of the HTTPHeaders object to add an HTTP header to your EWS call, as in the following example:
Change the Exchange object type that's associated with the service account. If the object is changed from a mailbox-enabled object to a mail user object, all EWS connections are routed to the back-end server that hosts the target mailbox.
ID članka: 2990048 - Poslednji pregled: 14.01.2015. - Verzija: 1