Looks like this bug was fixed in the trunk as part of http://
dev.zenoss.org/trac/ticket/1123
Also, there is a script in the trunk to remove local values of a
given zProperty from all devices in a given device class (and
optionally to recurse into sub device classes also.)
http://dev.zenoss.org/trac/browser/trunk/bin/zproprmlocal
http://dev.zenoss.org/trac/browser/trunk/Products/ZenUtils/
zproprmlocal.py
Instead of downloading those script files you can just execute the
following lines in zendmd to accomplish the same thing. Set devClass
and zPropName in those two lines to the device class and zProperty
name you wish to clear. This will operate recursively on all
subclasses of the device class you specify.
As always, it's a good idea to backup your data before making wide-
ranging changes. Tar/zip your entire $ZENHOME and put it someplace
safe or use zenbackup.
jasons-macbook:/usr/local/zenoss/bin jason$ zendmd
Welcome to zenoss dmd command shell!
use zhelp() to list commands
>>> from Acquisition import aq_base
>>> devClass = '/'
>>> zPropName = 'zSnmpVer'
>>> org = dmd.Devices.getOrganizer(devClass)
>>> for d in org.getSubDevicesGen():
... if hasattr(aq_base(d), zPropName): # 2 spaces at beginning of
this line
... d._delProperty(zPropName) # 4 spaces at beginning of this line
... # Just hit return here
>>> commit()
This code is deleting zProperty values from Devices to that they
inherit from their Device Class. Be careful doing this, you could
potentially screw up some of your devices if you use this on the
wrong class/zProperty. You have been warned :-)
-jason
On Mar 27, 2007, at 9:17 AM, Jason Stevens wrote:
I've created a ticket for this issue: http://dev.zenoss.org/trac/
ticket/1128
In addition to correcting the bug we'll throw together a script to
avoid the pain of manually deleting the local value of zSnmpVers
for devices already modeled. Stay tuned.
-jason
On Mar 24, 2007, at 11:15 PM, Greg Davis wrote:
Sorry, typo.
New devices are set to snmp v1 by default instead of inheriting
the zProp from the parent container.
--Greg
Greg Davis wrote:
Just to chime in here...
I have had this issue for a while, new devices are set to snmp
v2c by default instead of inheriting the zProp from the parent
container. I use ONLY v2c but i always have to delete the local
prop from new devices. Furthermore some function in Zenoss (not
sure but I think it may be zenmodeler) resets the snmp version
to 1. From time to time I get a device that is working normally
but generating all kinds of events because some of the monitored
OIDs are v2c only. I go delete the local SnmpVer zProperty and
everything sorts itself out. Can we look into why this important
setting is not handled correctly?
--Gregory Davis
Erik A. Dahl wrote:
Daniel,
You actually need to delete the property using the popup all the
way at the bottom of the zProperties page called "Delete Local
Property". Blanking out the field won't do it.
-EAD
On Mar 23, 2007, at 10:47 AM, Daniel McKinney wrote:
Erik,
Ok, I went to the following link: http://192.168.204.29:8080/
zport/dmd/Devices/Server/Linux/32-bit/devices/www3/zPropertyEdit
Then i scrolled down to zSnmpVer and emptied out that field.
Then hit Save.
I was expecting it to be populated with the v2c from the parent
zSnmpVer field set at this link: http://192.168.204.29:8080/
zport/dmd/Devices/zPropertyEdit
It wasn't. It was still blank. I thought maybe it needed to
have the collection collected using the Manage tab. Still no
luck.
I went so far as stopping all of Zenoss then starting it back
up and checking that page again. It's still blank.
Am i doing something wrong?
Thanks,
-Daniel
Erik A. Dahl wrote:
Daniel,
In stead of changing it on the device you need to delete its
local zProperty. This will allow the higher level one to set
the value. I think the problem is the local property
overriding the one higher up. You can delete the property at
the bottom of the zProperties page.
-EAD
On Mar 22, 2007, at 9:00 AM, Daniel McKinney wrote:
Erik,
Not sure what you mean by bad device. Do you mean the
devices that are complaining about bad OIDs? The only time i
get bad OID messages (after i researched some more) is when i
move a device from one Monitor to my new custom one. When
that occurs, the zSnmpVer field gets reset to v1 and that is
when the bad OIDs start.
Other than that, i still get the same problem Rob is having.
I go to a Class, set the zProperties' zSnmpVer to v2c and i
still have to manually go to each device beneath that Class
and change the zSnmpVer to v2c.
Thanks,
-Daniel
Erik A. Dahl wrote:
Hum,
Can you look on a bad device and see if the version has been
defined _just_ for that device? I wonder if the modeler is
messing this up. Shouldn't be.
-EAd
On Mar 21, 2007, at 7:47 AM, Daniel McKinney wrote:
Rob,
I had the same problem and i couldnt find a solution.
Actually, to make things worse, sometimes after i make the
change, it periodically reverts itself back to v1 and i
cant find out why just yet.
I originally thought it was reverting certain devices back
to v1 after i did an ./zenperfsnmp run -v10 -d
<device_name>. But i cant duplicate it. I basically do my
normal tinkering and anytime i get alot of "BAD OID" events/
messages in the zenperfsnmp.log file i check the zSnmpVer
and sure enough, its set back to v1. Very frustrating.
I think i am going to send a separate email to the zenoss
dev mailing list asking as a feature request to be able to
toggle the Snmp Version at the "Add Device" phase.
Does anyone know where the default setting is for the
zSnmpVer field? If i could set the default zSnmpVer to
v2c, i would be in business.
Before anyone says it, the very first step i did was click
on Devices/zProperties, went down to the zSnmpVer field and
set that to v2c a long time ago, so that cant be where the
"default" setting is. I know this because if i add a
device, it always adds it with v1 and i have to go in
afterwards and change it to v2c every time.
Thanks,
-Daniel
rgrijalvo wrote:
Hi All,
I'm hoping my logic is correct, but if someone could
verify Zenoss' behavior for the following, I'd really
appreciate it.
I was under the impression that if I changed any of the
zProperties for a certain class, in this case, /Device/
Network/Router, the changes would apply for all devices
within the class.
The parameter I am having issues with is zSnmpVer. Under
the class router, I changed it to v2. The path for this
parameter is listed as /Network/Router. However, when I
go to the zProperties tab of a newly discovered device of
class Router, the zSnmpVer value is back to the default of
v1. It does show a new path of /Network/Router/devices/
<device_name>. The new path does not really surprise me,
but I thought it would use what I defined at the Router
class level.
Am I correct in my expectations? I modified the
zCollectorCollectPlugins parameter to exclude the RouteMap
plugin and it propagated to all devices discovered with
that class.
Thanks everyone for their feedback.
Regards,
Rob-
-
-------------------- m2f --------------------
Read this topic online here:
http://community.zenoss.com/forums/viewtopic.php?p=4839#4839
-------------------- m2f --------------------
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
