| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 1904.1 | A couple of answers | NETRIX::"myrdal@zk3.dec.com" | Gregory P. Myrdal | Mon Mar 03 1997 13:43 | 17 |
| Jun-Mok,
1. Modify your network set in the asemgr utility (ie. run the a) and
then n) options). ASE does not detect that this interface can not
talk and operate on the state change by default. You need to add
this interface (and corresponding ones for all members) to the ASE
network onfiguration as a monitored interface. When all monitored
interfaces fail, the ASE services fail over.
2. These appear to be the result of running the cluster monitor. If
you are running cmon, this would explain them. The cluster monitor
runs the asemgr command to get the state of the ASE for its display.
Regards,
-- Greg
[Posted by WWW Notes gateway]
|
| 1904.2 | more question | DEKVC::JUNMOKSONG | | Tue Mar 04 1997 06:27 | 18 |
| Hi,
Thank you for your answer.
We have already configurated the FDDI monitored but this error
happened. I now knew the ifconfig command was set in start script
by "ifconfig alias fta0 alias shareloginname netmask 255.255.255.0"
and so we will set this by "/var/ase/sbin/nfs_ifconfig $$
start ${svcName} returnValue=$?"
What do you think of this?
And I will try ty cmon and can I get the documentation about it?
Best Regards,
Jun-Mok Song
|
| 1904.3 | This test case should work | NETRIX::"myrdal@zk3.dec.com" | Gregory P. Myrdal | Tue Mar 04 1997 14:59 | 25 |
| and so we will set this by "/var/ase/sbin/nfs_ifconfig $$
start ${svcName} returnValue=$?"
This looks ok, I created a login service as used the following
in my start script:
/var/ase/sbin/nfs_ifconfig $$ start $svcName
exit $?
Unfortunately I do not think it will fix your problem. Your ifconfig
line looked ok. I set up a login service and your example worked for
my case. The only problem is, I did not have fddi to test it out on. I
had to use ethernet. This should work, if it does not for you a QAR should
be filed.
And I will try ty cmon and can I get the documentation about it?
Try the Cluster Monitor manual pages:
http://www.zk3.dec.com/~buttlar/cluster/AS_ADMIN/CHPTR001.HTM#BLOCK45
-- Greg
[Posted by WWW Notes gateway]
|
| 1904.4 | More question | DEKVC::JUNMOKSONG | | Tue Mar 04 1997 22:23 | 16 |
|
Hi,
Thank you for your answer.
I have more question about "all monitored interface fail".
What does it mean? In my case, we had two MCs and one FDDI
and so I wonder if all monitored interface contain the two
MCs and one FDDI and ans consequently you mean only if two
MCs and one FDDI are disconnected or failed the login service
can be failover in Trucluster.
Thank you for cm message WEB site.
Best Regards,
Jun-Mok Song
|
| 1904.5 | We relocate services when all monitored interfaces fail | NETRIX::"myrdal@zk3.dec.com" | Gregory P. Myrdal | Wed Mar 05 1997 17:43 | 21 |
| I have more question about "all monitored interface fail".
What does it mean? In my case, we had two MCs and one FDDI
and so I wonder if all monitored interface contain the two
MCs and one FDDI and ans consequently you mean only if two
MCs and one FDDI are disconnected or failed the login service
can be failover in Trucluster.
A monitored interface is one in which you set to be monitored
in the asemgr network modification menu. By default we do not
set the MC networks as being monitored. Since clients do not
hang off of this "network" it is not useful to determine what do
do with services if one of them fails. In your case I would only
set the one fddi as being monitored.
Yes, the entire set of monitored interfaces have to fail in order for
services to be relocated.
Cheers,
-- Greg
[Posted by WWW Notes gateway]
|