Todd,
Something whacky is going on here and I don't have any great ideas as
to what it is. So let's do a little fishing..
First of all, try the zedmd script I gave you before with a small
change that I should have included originally (note the addition of
aq_base in the conditional):
$ zendmd
>>> from Acquisition import aq_base
>>> for t in dmd.Devices.getAllRRDTemplates():
... if not hasattr(aq_base(t), 'graphDefs'):
... t.buildRelations()
... print t.getPrimaryId()
Next, try navigating to one of your broken templates via the zmi.
Click on a link for a broken template, then add /manage_main to the
end of the url. You should see the oh-so-styling zope management
interface with a list of attributes on that template. "graphDefs"
should be one of the 6 or so items in that list.
Then, try to bring up the template in zendmd and see if has graphDefs
defined. The exact path you need to type depends on what template
you are using, but it should look something like this:
$ zendmd
>>> t =
dmd.Devices.SomeDeviceClass.MaybeSomeSubClass.rrdTemplates._getOb
('YourTemplateId')
>>> t
<RRDTemplate at /zport/dmd/Devices/SomeDeviceClass/maybeSomeSubClass/
rrdTemplates/YourTemplateId>
>>> t.graphDefs()
Hopefully the last line will result in some nice list, probably
empty. Maybe you'll get the same AttributeError that you've been
seeing on the zenoss pages. If you see that error then try this:
>>> t.buildRelations()
>>> t.graphDefs()
If you still get an error from t.graphDefs() then things are in a
very sad state. If that works then things are in a slightly less sad
state. Let me know and we'll continue from there.
One other thing to check for - make sure you only have one instance
of zeo running on your zenoss server. If you do a zenoss stop then
ps -eaf | grep python you'll get a hopefully short list that
shouldn't include anything with zenoss or zope or zeo in the
command. Also, try deleting all the zeo cache files while you have
zenoss stopped:
$ rm $ZENHOME/var/*.zec
-jason
On Oct 11, 2007, at 11:06 AM, Todd Davis wrote:
Upgraded to 2.0.93 -- this still occurs. Additionally, if I try to
edit ANY
of the templates I get the same error. However, I can add and edit
a new
template.
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Todd Davis
Sent: Wednesday, October 10, 2007 9:26 PM
To: 'General discussion of using zenoss system'
Subject: RE: [zenoss-users] Another 2.0.92 data issue
Didn't seem to recurse. I took out the if statement to just get a
list, but
didn't get all of them. I definitely noticed one template that is
set at an
individual device that wasn't in the list.
Also, it doesn't appear to be just filesystem graphs. I get it on
all SNMP
based graphs.
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Jason Stevens
Sent: Wednesday, October 10, 2007 5:41 PM
To: General discussion of using zenoss system
Subject: Re: [zenoss-users] Another 2.0.92 data issue
The call to getAllRRDTemplates() recurses through all device classes,
devices and device components, so it should have hit the template
you're having trouble with. Are you able to navigate to the
appropriate FileSystem template itself in zenoss? Are any graphs
listed there?
-jason
On Oct 10, 2007, at 6:02 PM, Todd Davis wrote:
Odd. I ran the for loop in dmd. No results. Evidently, all of
those
top-level templates have that attribute.
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Jason Stevens
Sent: Wednesday, October 10, 2007 2:53 PM
To: General discussion of using zenoss system
Subject: Re: [zenoss-users] Another 2.0.92 data issue
Todd,
If you're comfortable with zendmd I'd be interested to see the output
from this:
$ zendmd
for t in dmd.Devices.getAllRRDTemplates():
... if not hasattr(t, 'graphDefs'): # this line indented 2 spaces
... print t.getPrimaryId() # this line indented 4 spaces
This will print out a list of RRDTemplates that don't have the
relationship that appears from your error message to be missing. You
might be able to fix the problem you're seeing by adding this line to
the command above:
... t.buildRelations() # this line indented 4 spaces
-jason
On Oct 10, 2007, at 1:37 PM, Todd Davis wrote:
I upgraded from 2.0.90. I've periodically had problems with the
migrate
script(s). I don't remember if that one specifically had problems.
I've re-run the two you specified. They appear to complete
successfully.
Restarted Zenoss, but I get the same error if I go into one of the
graphs.
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Jason Stevens
Sent: Wednesday, October 10, 2007 8:37 AM
To: General discussion of using zenoss system
Subject: Re: [zenoss-users] Another 2.0.92 data issue
Todd,
As for the AttributeError:graphDefs problem, it looks like one of
the
data migrate scripts did not complete properly. Give this a try:
zenmigrate --step Thresholds
zenmigrate --step GraphDefinitionsAndFriends
Which version of zenoss did you upgrade to 2.0.92 from?
-jason
From: "Todd Davis" <[EMAIL PROTECTED]>
Date: October 9, 2007 4:48:45 PM EDT
To: "'General discussion of using zenoss system'" <zenoss-
[EMAIL PROTECTED]>
Subject: [zenoss-users] Another 2.0.92 data issue
Reply-To: General discussion of using zenoss system <zenoss-
[EMAIL PROTECTED]>
Here's another issue I'm having in 2.0.92.
Trying to view filesystem graphs gives the following:
* Module Products.PageTemplates.ZRPythonExpr, line 47, in
__call__
__traceback_info__: here.getDefaultGraphDefs(drange=drange)
* Module Python expression "here.getDefaultGraphDefs
(drange=drange)",
line 1, in <expression>
* Module Products.ZenModel.RRDView, line 202, in
getDefaultGraphDefs
* Module Products.ZenModel.RRDTemplate, line 117, in
getGraphDefs
AttributeError: graphDefs
Also, zenperfsnmp doesn't appear to be gathering information
correctly
(filesystems show no usage). Zenperfsnmp shows the following:
2007-10-09 15:41:27 INFO zen.zenperfsnmp: fetching default
RRDCreateCommand
2007-10-09 15:41:27 INFO zen.zenperfsnmp: fetching snmp status
2007-10-09 15:41:27 INFO zen.zenperfsnmp: Initiating incremental
device load
2007-10-09 15:41:27 ERROR zen.zenperfsnmp: Error loading devices:
[Failure
instance: Traceback (failure with no frames): exceptions.TypeError:
iteration over non-sequence
]
2007-10-09 15:41:27 INFO zen.zenperfsnmp: Count 2 good 0 bad 0 time
0.000060
2007-10-09 15:41:27 INFO zen.zenperfsnmp: Count 2 good 1 bad 0 time
0.014323
I wonder what that TypeError might be.
_______________________________________________
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
