This bug is related to 3505, the crash was fixed. Opening this new bug to track the hang. Refer to bug 765237 for background info

it looks like my pic code hangs a very long time...more than 3-4min to write a output file with 77MB....with the "old" stripe.c file this was the position glusterfsd crashed....now it didn't crash anymore, but the whole code hangs over two minutes....and it will become even worse after the first hang...

Time  0.158128818776E-12 and iteration    470 after    166.4 seconds
Time  0.161493261729E-12 and iteration    480 after    169.9 seconds
Time  0.164857704681E-12 and iteration    490 after    173.3 seconds
Time  0.168222147634E-12 and iteration    500 after    176.7 seconds
Time  0.171586590587E-12 and iteration    510 after    180.2 seconds
Time  0.174951033539E-12 and iteration    520 after    183.7 seconds
Time  0.178315476492E-12 and iteration    530 after    187.4 seconds
Time  0.181679919445E-12 and iteration    540 after    191.0 seconds
Time  0.185044362397E-12 and iteration    550 after    194.8 seconds
Time  0.188408805350E-12 and iteration    560 after    198.4 seconds

until here everthing is fine, serveral output files were written succesfully and also very fast! (in the old code nothing crashed until here so far!)

Time  0.191773248303E-12 and iteration    570 after    202.0 seconds
Time  0.195137691255E-12 and iteration    580 after    205.7 seconds <-- after this timestep, glusterfsd typically crashed....now: no crash but very long time for the next time-step
Time  0.198502134208E-12 and iteration    590 after    430.2 seconds
Time  0.201866577161E-12 and iteration    600 after    443.8 seconds
Time  0.205231020113E-12 and iteration    610 after    694.0 seconds <-- same probleme here, long time for writing a file