> Cliff White <[email protected]> wrote:
> >
> >
> > Tried hard to test this, but all it does for me is panic.
> > Is this fixed in 2.5.63?
> > This is 4-way PIII system.
> > panic, while booting
> > Press Y within 1 seconds to force file system integrity check...
> > [<c02409e8>] as_next_request+0x38/0x50
>
> We have some rough edges in the anticipatory scheduler. Please
> boot with elevator=deadline for now.
>
That was the ticket. Runs DBT2 fine now.
Detailed results for 2.5.62-mm3 w/plm-1592 (flock)
http://www.osdl.org/archive/cliffw/flock/mm3/index.html
I didn't have profiling turned on for this run. will be doing
a few more runs to get profile data.
The results are a very tiny bit faster than 2.5.62 stock.
(TPM change < 10% )
VMstat looks a bit different. a quick snap below.
2.5.62 - stock (+ plm-1592)
procs memory swap io system cpu
r b w swpd free buff cache si so bi bo in cs us sy id
5 0 0 280 151340 29292 1396592 0 0 159 12443 1891 5497 95 4 1
4 0 0 280 145564 29536 1396976 0 0 193 1000 1149 3412 95 2 3
5 1 0 280 136972 29884 1397376 0 0 192 1036 1153 3554 95 2 3
5 0 0 280 129044 30116 1397764 0 0 178 1042 1152 3479 95 2 3
5 0 0 280 122260 30344 1398140 0 0 153 1011 1145 3356 95 2 3
5 1 0 280 114968 30576 1398520 0 0 153 1023 1146 3426 95 2 3
4 0 0 280 108252 30808 1398908 0 0 149 1007 1144 3340 95 2 3
4 0 0 280 101416 31140 1399300 0 0 133 1038 1147 3444 95 2 3
2.5.62-mm3 (+ plm-1592)
procs memory swap io system cpu
r b w swpd free buff cache si so bi bo in cs us sy id
7 1 0 980 10296 25656 1528952 0 0 207 1123 1259 3741 94 3 3
6 2 0 980 9592 25752 1521964 0 0 190 1116 1256 3673 94 3 3
5 2 0 980 9848 25912 1514212 0 0 177 1117 1255 3641 94 3 3
6 1 0 980 13368 26040 1504004 0 0 157 1128 1253 3779 94 3 3
6 1 0 980 9912 26164 1501048 0 0 145 1107 1249 3606 94 3 3
6 1 0 980 9656 26244 1494832 0 0 151 1134 1253 3721 95 3 2
7 2 0 980 9896 26296 1487988 0 0 144 1139 1253 3761 94 3 3
12 9 0 980 10176 26348 1480312 0 0 103 12019 1986 5867 94 5 1
----------------------
cliffw