Thanks for a bug report. Could you install debugging information for evolution-data-server and evolution, then get evolution into this state and then capture a backtrace of it, to see what it does, please? It can be that it is waiting for a response from the server, where is set a 90 seconds timeout, but it depend son couple aspects, like what account type you've setup and so on.You can get the backtrace with command like this:
   $ gdb --batch --ex "t a a bt" -pid=`pidof evolution` &>bt.txt
Please check the bt.txt for any private information, like passwords, email address, server addresses,... I usually search for "pass" at least (quotes for clarity only).

I haven't installed debug information, but did run GDB:

I hope this can be useful nevertheless.

-----------------

[New LWP 2873]
[New LWP 2844]
[New LWP 2841]
[New LWP 2840]
[New LWP 2839]
[New LWP 2838]
[New LWP 2837]
[New LWP 2836]
[New LWP 2835]
[New LWP 2834]
[New LWP 2812]
[New LWP 2811]
[New LWP 2810]
[New LWP 2809]
[New LWP 2808]
[New LWP 2807]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/i386-linux-gnu/i686/cmov/libthread_db.so.1".
0xb7746be0 in __kernel_vsyscall ()

+ Trace 236202

Thanks for the update. Without the debuginfo it's significantly harder to know what the code does, due to too many "question marks" in the backtrace, though some functions are notated by name, thus one can at least guess.

I found a bug which contains a similar backtrace, with debuginfo installed, thus I'm marking this as a duplicate of it. Please see it for any further comments.

*** This bug has been marked as a duplicate of bug 742167 ***