1) The machine in question is doing quite a few context switches; around
1300-1400 per second in average, sometimes topping 4500 or more (as
reported by vmstat). This is a P2-266, and my question is: how many
context switches is it reasonable to expect from such a CPU? I realize
the answer depends a lot on the tasks the machine performs, any opinions
on the matter would be appreciated.
2) Is it possible to see how much time a single process spends waiting for
I/O reads and writes? I understand top on both solaris and HP/UX reports a
percentage measurement for this.
I know these questions are a bit off-topic and that everyone resents being
disturbed in their bashing of the only slightly esoterical issues of
virtual memory management algorithms, but I would appreciate any input in
the matter as this is a rather important project to us. We have invested a
lot of prestige in this project and it would greatly further the GNU/Linux
cause in Sweden if it became successful (as it is a rather high profile
project).
Please reply cc to me, as I am not normally on the kernel mailing list.
Quest
+------------------------------------------------------------------------+
| Three servers for the admins under the influence of rye, |
| seven routers for the network techs in their halls of stone |
| Nine workstations for mortal lusers doomed to die |
| One NT box from the dark lord on his throne |
| in the land of Redmond where the shadows lie |
| One box to run them all, one box to blind them |
| one box to control them all and in the darkness grind them |
| from the land of Redmond where the shadows lie. |
| --- Bram Smits / alt.sysadmin.recovery |
+------------------------------------------------------------------------+
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/