vacm-features Mailing List for VACM

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Hello,
	I have a brief question.

Does VACM run on any other hardware apart from the VALinux Boxes?

tvm
Zameer A.

Zameer -

As with almost everything - that depends.  Basically, the number of
nodes that can managed is dependent on the horsepower and load of the
machine that the nexxus is running on, and the mix of machines that
are being managed.  EMP-based machines like VA's 22XX and 45XX servers
require more resources on the nexxus server than machines like the
VA 1120/1220/1000.  You should be able to handle 12 nodes with a
typical server running the nexxus.  If the nexxus server is expected
to provide a lot of additional services, it will need additional
horsepower.

You can run multiple nexxus's under VACM, in order to scale up to
support additional nodes.  For example, you might want a nexxus per
rack full of 1U servers, and maybe share a nexxus among a couple of
racks of 4U's.  There is no hard-and-fast formula.

Jerry.

On Thu, Apr 12, 2001 at 11:26:51AM -0400, Ahmed, Zameer wrote:
> Hello,
> 	I am dabbling in setting clusters managed by VACM. I right now going
> on 12 nodes. Whats the maximum I can go with VACM.
> 
> TIA
> Zameer A.
> 
> _______________________________________________
> Vacm-features mailing list
> Vac...@li...
> http://lists.sourceforge.net/lists/listinfo/vacm-features

Hello,
	I am dabbling in setting clusters managed by VACM. I right now going
on 12 nodes. Whats the maximum I can go with VACM.

TIA
Zameer A.

On Tue, Aug 29, 2000 at 09:57:04AM +0530, Piyush Goswami wrote:
> Hello,
> 
> I have just started using vacm-2.0.0beta2 version .All the features seem
> 
> good to me.
> My requirement is a bit different.
> I am planning to have my web server replicated on another machine so
> that if one fails ,the other one can still serve the world the required
> web pages.
> Also when both machines are up, both should share the load between
> themselves to avoid CPU overloading .Is it possible with VACM?
> Does VACM provide such a feature so as to satisfy my needs?
> How do I go about it ?
> 
> Can anyone please suggest something?
> 
 Piyush.

Try <http://www.us.ultramonkey.org/>. It a load balancing/failover tool. 

Thanks,
John 

-----------------------------------------------------------------
John Goebel					VA Linux Systems
jg...@va...				408-621-7058
-----------------------------------------------------------------

Hello,

I have just started using vacm-2.0.0beta2 version .All the features seem

good to me.
My requirement is a bit different.
I am planning to have my web server replicated on another machine so
that if one fails ,the other one can still serve the world the required
web pages.
Also when both machines are up, both should share the load between
themselves to avoid CPU overloading .Is it possible with VACM?
Does VACM provide such a feature so as to satisfy my needs?
How do I go about it ?

Can anyone please suggest something?

Regards,

Piyush.

"Antonio W. Chan" wrote:
> 
> Hi,                                  May 2, 2000
> 
> I have been experimenting with VACM 2.0.0 beta release, which
> includes a GUI interface called hoover. It looks nice, and I
> would like to suggest some features for future versions.
> 
> 1) The "console" button doesn't seem to work properly. It
> launches a blank window, and then my hoover GUI freezes.
> I think the "console" button is not connecting to the
> correct tty port. Can this be fixed for the next version?
> 

Yes, this is something that I have been putting off, as I don't
have the right config yet to work with it. In fact, the latest
versions of hoover (CVS and SGI ACE) popup a dialog saying
that serial console functionality is deferred. In any case,
hoover shouldn't freeze.

> 2) My nexxus process sometimes dies, and a lot of hoover
> error messages appear on my console window because one of
> the monitored machines crashes OR is rebooted purposely
> without using hoover's "power down" or "reset" buttons.
> I think nexxus should handle that more gracefully.

I have been cleaning up the various console messages, but there
are still several left to do. I am trying to be manage the
messages carefully to not unduly annoy the user. For instance,
it would be horrible to popup a 'broken pipe' message for each
time that I get the error condition. An example of that
management is the failure to connect with a nexxus in beginning
popups up the error dialog and keeps the user apprised about its
attempts to connect.

> 
> 3) The "reset" button in hoover should be relabeled
> "reboot"
> 

What do other people think about that? I think it is a good
idea, but its referred to as 'reset' everywhere else in 
EMP-land.

> 4) Hoover should be on "continuous-update" mode when
> running. When I powered-down one of the monitored nodes,
> the status of that node on the hoover GUI still said
> "RUNNING". This change would make it a more powerful
> tool.

It is in continuous update mode and rechecks all the nodes every
minute, unless the node is down or negotiating (or other 'odd' 
condition), in which case it rechecks every 15 seconds. You can make 
it recheck more frequently with the '-i <mins>' argument, such as '-i
0.25' 
will check every 15 seconds. If you are still not getting things updated 
correctly, please send me mail (dt...@fu...) and we can 
investigate more.

> 
> 5) Hoover should also provide another button called
> "alarm-notify", with certain configuration options
> (levels of alarms, send e-mail on alarm, bell on alarm,
> send a signal to a beeper on alarm, etc) for the user
> to choose. Again, this would make VACM a more powerful
> monitoring tool.
> 

These things are going to crop in a number of different places, not
just hoover. San has some wonderful plans for such functionality and
I intend on adding heuristics to hoover to coalesce diagnostic messages
into more useful information.

Thanks for the feedback and please feel free to directly send me any
further
feedback you have about hoover.

	-Dean Johnson

Hi,                                  May 2, 2000

I have been experimenting with VACM 2.0.0 beta release, which
includes a GUI interface called hoover. It looks nice, and I
would like to suggest some features for future versions.

1) The "console" button doesn't seem to work properly. It
launches a blank window, and then my hoover GUI freezes.
I think the "console" button is not connecting to the 
correct tty port. Can this be fixed for the next version?

2) My nexxus process sometimes dies, and a lot of hoover
error messages appear on my console window because one of
the monitored machines crashes OR is rebooted purposely
without using hoover's "power down" or "reset" buttons.
I think nexxus should handle that more gracefully.

3) The "reset" button in hoover should be relabeled
"reboot"

4) Hoover should be on "continuous-update" mode when
running. When I powered-down one of the monitored nodes,
the status of that node on the hoover GUI still said
"RUNNING". This change would make it a more powerful 
tool.

5) Hoover should also provide another button called
"alarm-notify", with certain configuration options
(levels of alarms, send e-mail on alarm, bell on alarm, 
send a signal to a beeper on alarm, etc) for the user 
to choose. Again, this would make VACM a more powerful
monitoring tool. 

Cheers,

Tony

On Tue, Apr 25, 2000 at 12:54:26PM -0700, Dale Harris wrote:
> Something else that might be nice for sercon_terminal is an escape sequence,
> where we could at least terminate the session, or (if you want to get 
> really fancy) back to a prompt like telnet.

sercon_terminal is just a test application. its not designed to be used by
actual end users. if someone wants to write a real console terminal program
then by all means, do so :) It's just designed to test the module and show
how one would interface with the module
-san

-- 
-----------------------------------------------
San Mehat
Senior Software Engineer & Digital Groovemaster
VA Linux Systems
net...@va...
408-542-8654
-----------------------------------------------

Something else that might be nice for sercon_terminal is an escape sequence,
where we could at least terminate the session, or (if you want to get 
really fancy) back to a prompt like telnet.

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dale Harris <ro...@va...>                         VA Linux Systems
Sr. Support Engr.                                        1382 Bordeaux Dr.
(408) 543-8025                                           Sunnyvale, CA 94089

2000	Jan	Feb	Mar	Apr (2)	May (2)	Jun	Jul	Aug (2)	Sep	Oct	Nov	Dec
2001	Jan	Feb	Mar	Apr (2)	May	Jun (1)	Jul	Aug	Sep	Oct	Nov	Dec