It's because the workload has shifted to scattered writes, since under the glitz it's 'only' a 7-drive RAID-6 array, so it has very limited sustained random write performance.  Notice also that the write workload shows up as *read* latency due to the caching.

Well now almost all your workload is writes!  Can you post the disk load info (fig.7 on pg.16 in that doc)?

Im not really sure that i understand the concept of scattered writes :-)

I looked through the following document, that talks about especially VDI projects, on hybrid arrays:

en.community.dell.com/.../2642.sizing-and-best-practices-for-deploying-vmware-view-vdi-with-equallogic-storage-by-sis.aspx

the IO pattern of XenApp and VDI should be very similar.

Admittedly they do not show the Full graphs in the report/commercial.

Since my first post actually included a screenshot of mostly IO during the night, i have added a new picture below, that shows the IO over a day. 8 in the morning to 16

Sure can.

sorry im a bit limited on the resolution of my laptop, so i had to cut a few of the harddrives out, but the IO on them is the same more or less.

That is odd, although the throughput had dropped by then none of those drives look more than 5 to 10% loaded.  Have you opened a support case?

No i kinda wanted to hear peoples opinion before wasting support time on it :)

Don't be afraid to use support - you've paid for it!

I know, but my time is limited ;)

I should open a support case, you're disk queue depth is way to high