wow, I can't really type this in as normal! I'll type it in again but duck!
woooooow I caan'''tt rrrrreeeeeaaaalllyyy tttttyppppeeee tthiiiss inn aaaaaaaas noormmmmaaaaaalllll!!!!!!
::sigh:: and I don't even drink!
I have read spots where it could be problems w/the kernel and AMD's newer tech "Cool and Quiet". I really dunno which way to take w/this
thx
John Rose
On Fri, 2006-09-22 at 19:19, rado wrote:
wow, I can't really type this in as normal! I'll type it in again but duck!
woooooow I caan'''tt rrrrreeeeeaaaalllyyy tttttyppppeeee tthiiiss inn aaaaaaaas noormmmmaaaaaalllll!!!!!!
http://kb.vmware.com/vmtnkb/search.do?cmd=displayKC&docType=kc&exter... and this for the clock speed
http://kb.vmware.com/vmtnkb/search.do?cmd=displayKC&docType=kc&exter...
::sigh:: and I don't even drink!
I have read spots where it could be problems w/the kernel and AMD's newer tech "Cool and Quiet". I really dunno which way to take w/this
An alternative would be to enable XDCMP logins and use Xnest to connect to the virtual machines directly instead of using the console program. Likewise you can use rdesktop to connect to the terminal services (desktop sharing on XP) on windows boxes.
On Fri, 2006-09-22 at 22:10 -0500, Les Mikesell wrote:
On Fri, 2006-09-22 at 19:19, rado wrote:
wow, I can't really type this in as normal! I'll type it in again but duck!
woooooow I caan'''tt rrrrreeeeeaaaalllyyy tttttyppppeeee tthiiiss inn aaaaaaaas noormmmmaaaaaalllll!!!!!!
http://kb.vmware.com/vmtnkb/search.do?cmd=displayKC&docType=kc&exter... and this for the clock speed
http://kb.vmware.com/vmtnkb/search.do?cmd=displayKC&docType=kc&exter...
::sigh:: and I don't even drink!
I have read spots where it could be problems w/the kernel and AMD's newer tech "Cool and Quiet". I really dunno which way to take w/this
An alternative would be to enable XDCMP logins and use Xnest to connect to the virtual machines directly instead of using the console program. Likewise you can use rdesktop to connect to the terminal services (desktop sharing on XP) on windows boxes.
Saving these little gems, although both problems are seemingly 1/2 assed under control now...we shall see...btw this is normal typing.
last nite I rebooted and in cmos actuall enabled typmatic rate to it's minimums so it so maybe it's actually making it mind.
right now system times are off 2 mins after all nite ...yest. that would have been 10-12hrs. I stopped all ntp or whatever that is...going out to one of the time servers...just getting time from the host system.
I have no clue if these 2 tweaks will hold over time but It did have all nite to gain time as it was doing and didn't
thx
John
rado wrote:
right now system times are off 2 mins after all nite ...yest. that would have been 10-12hrs. I stopped all ntp or whatever that is...going out to one of the time servers...just getting time from the host system.
Some hints from the vmware website for 2.6 kernels say to use "clock=pit" kernel option in guests, disable NTP in guests (system clocks in virtual machines are too unstable), and let vmware tools sync the time with the host OS (you can use NTP to sync the time of the host OS with some time source, of course).
If you have both 2.4 and 2.6 kernels running in guests, you'll probably notice that 2.6 kernels are taking 3-4 times more CPU time when the guest is idling. That's because the default frequency of interrupts changed from 100Hz to 1kHz (and that is per CPU) in 2.6 kernels. Placing much more stress on vmware and causing all bunch of problems with it. It's configurable compile time option (you'd need to recompile kernel to change it to the old default). I have one (not really brand new anymore) dual-Xeon machine that runs ESX with 5 2.6 guests and couple of 2.4 guests. When all the guests are idle, they consume about half of the CPU time of one processor. I haven't yet attempted recompiling 2.6 kernel and lowering frequency of interrupts, but it's probably worth trying.
On Sat, 2006-09-23 at 07:26 -0500, Aleksandar Milivojevic wrote:
rado wrote:
right now system times are off 2 mins after all nite ...yest. that would have been 10-12hrs. I stopped all ntp or whatever that is...going out to one of the time servers...just getting time from the host system.
Some hints from the vmware website for 2.6 kernels say to use "clock=pit" kernel option in guests, disable NTP in guests (system clocks in virtual machines are too unstable), and let vmware tools sync the time with the host OS (you can use NTP to sync the time of the host OS with some time source, of course).
If you have both 2.4 and 2.6 kernels running in guests, you'll probably notice that 2.6 kernels are taking 3-4 times more CPU time when the guest is idling. That's because the default frequency of interrupts changed from 100Hz to 1kHz (and that is per CPU) in 2.6 kernels. Placing much more stress on vmware and causing all bunch of problems with it. It's configurable compile time option (you'd need to recompile kernel to change it to the old default). I have one (not really brand new anymore) dual-Xeon machine that runs ESX with 5 2.6 guests and couple of 2.4 guests. When all the guests are idle, they consume about half of the CPU time of one processor. I haven't yet attempted recompiling 2.6 kernel and lowering frequency of interrupts, but it's probably worth trying.
Hi, Alex, thx for this..when I get time, I will look into this.
another interesting thing is that I guess I'm still having this grrrr key repeat going oonnnnnnnnnnnnnn lloolll
jr
-
Aleksandar Milivojevic -
Les Mikesell -
rado