FC 6.1 Hangs Windows 7 Topic is solved

[Support]

mmm...so this issue might be related to the Level-2 caching...
To be more correct, could you test more time under the current FC settings? If everything runs smoothly, then you might try the level-2 caching with MBU L2 algorithm. I remember that previously you used LBW algorithm. I am not sure if it is the cause, but we'll see the results.
Thanks.

[Michail]

Now it's 2 hours without hangs, I'll enable Def.Write back to see if it give some difference, test it today, and tomorrow I'll enable L2 MBU

[Support]

Since you didn't enable Defer-Write during previous hangs, I think it is not necessary to test with Defer-Write feature.

[Michail]

I agree, but who knows, maybe there is some other incompatibility... If not... then it's even better

[Support]

ok, we'll wait for the results.
PS. please don't enable Defer-Write while testing L2 caching, otherwise the results might be confusing.

[Michail]

Maybe then it's even better to start with Read-Only Caching Strategy with MBU L2? I think it'll narrow the search for the problem source, and will give less stress to my data on hangs

[Support]

Read-Write strategy is ok. Thus we keep almost same conditions. Otherwise we might mix the results.
Later we'll test other factors if it can't get the cause.

[Michail]

Ok. I've tired of waiting hangs with L1 only cache and enabled L2

FC_L1_L2_RW.jpg (152.29 KiB) Viewed 5160 times

After that rebooted, started Perf.Stat.Monitor, started Thunderbird, Opera, Firefox, reloaded 3 pages in FF and in about 3 minutes after restarting get hang

[Michail]

Maybe to test with L2 on drive D? Maybe problem comes from SSD itself? (But for caching drive D it worked OK...)

[Support]

ok...Thank you, Michail.
Can you help to check the "LowerFilters" and "UpperFilters" values in the below branch in the registry?
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Class\{71A27CDD-812A-11D0-BEC7-08002BE2092F}

We'll trace this issue and let you know if there is any updates. I think the biggest problem is that we can't reproduce the issue, but anyway we'll check our code.