New build, fresh OS install.
This is a part-windows part-networking issue (I think). If this is more suited to Networks forum, I apologise.
Occasionally, seemingly at random, I will develop CPU spikes.
When it hits, I get a 2 second hang, followed by about 2 seconds of being able to actually move the mouse again, and then another pause. This will continue until I reboot. Monitoring the CPU clearly shows spikes of one core to 100%.
To find out more I ran process explorer and determined the cause of the spikes. Turns out to be Deferred Procedure Calls (DFCs). Now, after a bit of research I found I could monitor these with RATTV3.
Took me a while to digest what the logs from RATTV3 showed but I think I found the problem.
NDIS.sys seems to be the cause of the DPC spikes if I'm correct, since those times are far too high compared to all the others in the log. (I assume this although I'm not entirely sure I understand it.)
Now, secondly, I noticed the when the DPC spikes start, the wireless adaptor immediately cuts out. So, as far as I can tell the wireless adaptor is causing this? This make sense?
The adaptor is a Belkin F5D7000UK. I have been having problems on this rig with it dropping out completely, requiring me to "repair" it, before it would work again. This suggests there are issues. On my previous rig however, the card was fine and this issue did not come about. I have tried updating to the latest drivers and even tried the generic drivers supplied by RaLink however the problem seems to re-occur.
So, if anyone has any advice on this - I'd greatly appreciate it.
This is a part-windows part-networking issue (I think). If this is more suited to Networks forum, I apologise.
Occasionally, seemingly at random, I will develop CPU spikes.
When it hits, I get a 2 second hang, followed by about 2 seconds of being able to actually move the mouse again, and then another pause. This will continue until I reboot. Monitoring the CPU clearly shows spikes of one core to 100%.
To find out more I ran process explorer and determined the cause of the spikes. Turns out to be Deferred Procedure Calls (DFCs). Now, after a bit of research I found I could monitor these with RATTV3.
Took me a while to digest what the logs from RATTV3 showed but I think I found the problem.
Code:
Cumulative Histogram for: NDIS.sys (NDIS.sys)
#, Range, Label, ISR Count, DPC Count, DPCTmr Count
0, 0.00us to 0.99us, <1.00us, 0, 96371, 46773
1, 1.00us to 1.99us, <2.00us, 18641, 62289, 73064
2, 2.00us to 2.99us, <3.00us, 4820051, 370870, 69360
3, 3.00us to 3.99us, <4.00us, 1536593, 782946, 40438
4, 4.00us to 4.99us, <5.00us, 813596, 4065035, 10038
5, 5.00us to 5.99us, <6.00us, 70070, 411226, 4135
6, 6.00us to 6.99us, <7.00us, 12769, 291228, 1025
7, 7.00us to 7.99us, <8.00us, 4049, 319564, 1397
8, 8.00us to 8.99us, <9.00us, 2164, 149682, 1633
9, 9.00us to 9.99us, <10.00us, 809, 65472, 2653
10, 10.00us to 19.99us, <20.00us, 535, 574464, 20304
11, 20.00us to 29.99us, <30.00us, 2, 130055, 8
12, 30.00us to 39.99us, <40.00us, 0, 87345, 125
13, 40.00us to 49.99us, <50.00us, 1, 48237, 980
14, 50.00us to 59.99us, <60.00us, 1, 57121, 35
15, 60.00us to 69.99us, <70.00us, 0, 43291, 2
16, 70.00us to 79.99us, <80.00us, 0, 35595, 4
17, 80.00us to 89.99us, <90.00us, 0, 25973, 0
18, 90.00us to 99.99us, <100.00us, 0, 23418, 0
19, 100.00us to 109.99us, <110.00us, 0, 32152, 0
20, 110.00us to 119.99us, <120.00us, 0, 22362, 0
21, 120.00us to 129.99us, <130.00us, 0, 15856, 0
22, 130.00us to 139.99us, <140.00us, 0, 12421, 0
23, 140.00us to 149.99us, <150.00us, 0, 11721, 0
24, 150.00us to 159.99us, <160.00us, 0, 10490, 0
25, 160.00us to 169.99us, <170.00us, 0, 12437, 0
26, 170.00us to 179.99us, <180.00us, 0, 14880, 0
27, 180.00us to 189.99us, <190.00us, 0, 18776, 0
28, 190.00us to 199.99us, <200.00us, 0, 16737, 0
29, 200.00us to 209.99us, <210.00us, 0, 15229, 0
.......
66, 570.00us to 579.99us, <580.00us, 0, 1, 0
68, 590.00us to 599.99us, <600.00us, 0, 3, 0
71, 620.00us to 629.99us, <630.00us, 0, 1, 0
73, 640.00us to 649.99us, <650.00us, 0, 1, 0
74, 650.00us to 659.99us, <660.00us, 0, 1, 0
81, 720.00us to 729.99us, <730.00us, 0, 1, 0
83, 740.00us to 749.99us, <750.00us, 0, 1, 0
89, 800.00us to 809.99us, <810.00us, 0, 1, 0
90, 810.00us to 819.99us, <820.00us, 0, 1, 0
96, 870.00us to 879.99us, <880.00us, 0, 1, 0
[b]1009, 10.00ms to 19.99ms, <20.00ms, 0, 592, 2
1010, 20.00ms to 29.99ms, <30.00ms, 0, 4, 0[/b]
1108, 1000.00ms and greater, >=1000.0ms, 0, 0, 263
TOTALS, 7279281, 7867531, 272239NDIS.sys seems to be the cause of the DPC spikes if I'm correct, since those times are far too high compared to all the others in the log. (I assume this although I'm not entirely sure I understand it.)
Now, secondly, I noticed the when the DPC spikes start, the wireless adaptor immediately cuts out. So, as far as I can tell the wireless adaptor is causing this? This make sense?
The adaptor is a Belkin F5D7000UK. I have been having problems on this rig with it dropping out completely, requiring me to "repair" it, before it would work again. This suggests there are issues. On my previous rig however, the card was fine and this issue did not come about. I have tried updating to the latest drivers and even tried the generic drivers supplied by RaLink however the problem seems to re-occur.
So, if anyone has any advice on this - I'd greatly appreciate it.
