headers
Matt Ray | 1 Jul 2010 03:03

Re: is zenoss optimized to use multicore procs?


Zenoss is actually a series of daemons, so while individual processes may not take full advantage of
multiple cores, the distinct individual processes will.
 
Thanks,
Matt Ray
Zenoss Community Manager
community.zenoss.org
mray <at> zenoss.com (mailto:mray <at> zenoss.com)

--------------------------------------------------------------

To post a reply to the thread message, either reply to this email (recommended) or visit the message page: http://community.zenoss.org/message/50183#50183

--end--
Permalink | Reply | 0 comments |
headers
hendra manto | 1 Jul 2010 06:10

Re: Different OID but graph is same


-F
-E
--height=100
--width=500
--vertical-label=Byte / Second
DEF:oid-01-raw=rrdPath/oid-01_oid-01.rrd:ds0:AVERAGE
DEF:oid-01-raw-max=rrdPath/oid-01_oid-01.rrd:ds0:MAX
CDEF:oid-01=oid-01-raw
AREA:oid-01-raw#00cc00ff:oid-01        
GPRINT:oid-01-raw:LAST:cur\:%5.2lf%s
GPRINT:oid-01-raw:AVERAGE:avg\:%5.2lf%s
GPRINT:oid-01-raw-max:MAX:max\:%5.2lf%s\j
 
 
-F
-E
--height=100
--width=500
--vertical-label=Byte / Second
DEF:oid-03-raw=rrdPath/oid-03_oid-03.rrd:ds0:AVERAGE
DEF:oid-03-raw-max=rrdPath/oid-03_oid-03.rrd:ds0:MAX
CDEF:oid-03=oid-03-raw
AREA:oid-03-raw#7B3F00ff:oid-03        
GPRINT:oid-03-raw:LAST:cur\:%5.2lf%s
GPRINT:oid-03-raw:AVERAGE:avg\:%5.2lf%s
GPRINT:oid-03-raw-max:MAX:max\:%5.2lf%s\j

--------------------------------------------------------------
(Continue reading)

Permalink | Reply | 0 comments |
headers
carpej | 1 Jul 2010 14:34

Re: Loss of Multi-graph Reports w/new remote collector (updated)


ANyone else ran into this or have any insight into how multigraphs work that may need to be adjusted for a
remote collector? 

--------------------------------------------------------------

To post a reply to the thread message, either reply to this email (recommended) or visit the message page: http://community.zenoss.org/message/50218#50218

--end--
Permalink | Reply | 0 comments |
headers
jmp242 | 1 Jul 2010 14:47

Server migration, Zenhub won't stay up


Well, I've finally had to upgrade my Zenoss server hardware as the old server died. I took the opportunity to
upgrade to a 64 bit OS (EL5.4) and installed the 64 bit rpm then. I converted the RRD files using rrdtool dump
and restore... I then restored $ZENHOME and the databases. I can log in, the UI is fine (and much faster).
The events are all there, the configuration is all there.
 
However, none of the graphs are seen by Zenoss and don't show up in the UI (perf tab etc) - it claims it can't
find the files or they don't exist. The RRD files are named the same thing. I checked and the zenoss user has
read/write permissions on the directories.
 
Worse, zenhub won't stay up for some reason. This seems to also cause zenprocess to crash. I've turned
logging to debug but I don't get anything interesting in the zenhub log:
2010-07-01 08:04:23 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/Products/DataCollector/plugins
2010-07-01 08:04:23 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/Products/ZenWin/modeler/plugins
2010-07-01 08:04:23 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/Products/ApacheMonitor/modeler/plugins
2010-07-01 08:04:23 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/Products/HttpMonitor/modeler/plugins
2010-07-01 08:04:23 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/Products/ImprovedLinux/modeler/plugins
2010-07-01 08:04:23 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/Products/MySqlMonitor/modeler/plugins
2010-07-01 08:04:23 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/Products/WindowsWMIDeviceTemplateV2/modeler/plugins
2010-07-01 08:04:23 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/ZenPacks/ZenPacks.BlakeDrager.fping-1.0-py2.4.egg/ZenPacks/BlakeDrager/fping/modeler/plugins
2010-07-01 08:04:23 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/ZenPacks/ZenPacks.community.Perfmon-1.2-py2.4.egg/ZenPacks/community/Perfmon/modeler/plugins
2010-07-01 08:04:24 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/ZenPacks/ZenPacks.networking.HPProcurve-1.1-py2.4.egg/ZenPacks/networking/HPProcurve/modeler/plugins
2010-07-01 08:04:24 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/ZenPacks/ZenPacks.skills1st.bridge-1.0-py2.4.egg/ZenPacks/skills1st/bridge/modeler/plugins
2010-07-01 08:04:24 DEBUG zen.Plugins: Loading collector plugins from: /opt/zenoss/ZenPacks/ZenPacks.zenoss.NtpMonitor-2.0.2-py2.4.egg/ZenPacks/zenoss/NtpMonitor/modeler/plugins
2010-07-01 08:04:24 DEBUG zen.Events: Looking up IP localhost
2010-07-01 08:04:25 DEBUG zen.Events: EventClassInst=Start
2010-07-01 08:04:25 DEBUG zen.Events: Applying transform at /Events
2010-07-01 08:04:25 DEBUG zen.DbConnectionPool: Creating a new connection; Pool size: 0
2010-07-01 08:04:25 DEBUG zen.Events: update status set clearid =
'eef125c0-4b54-4c95-a7fb-21a6fc752691' where device='localhost' and component='zenhub' and
(Continue reading)

Permalink | Reply | 3 comments |
headers
car1 | 1 Jul 2010 15:14

Re: Error when accessing event console


After spending much time debugging the code I found the problem to be in the 'filteredWhere' method of the
EventManageBase class. This method takes as a parameter a dictionary in the variable 'filters'. This is
what was being passed in in my case:
 
filters = {'lastTime': '2010-06-29T00:00:00', 'firstTime': '2010-06-29T00:00:00', 'severity': [5,
4, 3, 2, 0], 'eventState': []}
 
Because the 'eventState' key had as a value an empty list, this piece of code created the "and ()" and
appended it to the query:
 
 
elif ftype=='multiselectmenu':
                if isinstance(v, basestring): v = (v,)
                sevstr = ' or '.join(['%s=%%s' % (k,) for s in v])
                queryValues.extend(v)
                newwhere += ' and (%s) ' % sevstr

 
 
So I add a test to see if sevstr is empty before appending it to the where clause:
 
 
elif ftype=='multiselectmenu':
                if isinstance(v, basestring): v = (v,)
                sevstr = ' or '.join(['%s=%%s' % (k,) for s in v])
                queryValues.extend(v)
                if sevstr:
                    newwhere += ' and (%s) ' % sevstr
(Continue reading)

Permalink | Reply | 0 comments |
headers
jmp242 | 1 Jul 2010 15:17

Re: Server migration, Zenhub won't stay up


To the extent that this may help my own question, an strace shows select queries just timing out so a
service mysqld status
gave the unhappy
 
mysqld dead but subsys locked
 
so it looks like it may all come down to MySQL not liking me for some odd reason.

--------------------------------------------------------------

To post a reply to the thread message, either reply to this email (recommended) or visit the message page: http://community.zenoss.org/message/50221#50221

--end--
Permalink | Reply | 0 comments |
headers
jmp242 | 1 Jul 2010 15:27

Re: Error when accessing event console


This certainly sounds like a bug and a ticket ought be created so that you don't have to make this change for
all future upgrades . . .
 
http://dev.zenoss.com/trac/wiki/HowToAddTicket
 
--
James Pulver
Information Technology Area Supervisor
LEPP Computer Group
Cornell University

--------------------------------------------------------------

To post a reply to the thread message, either reply to this email (recommended) or visit the message page: http://community.zenoss.org/message/50222#50222

--end--
Permalink | Reply | 0 comments |
headers
jmp242 | 1 Jul 2010 15:27

Re: Server migration, Zenhub won't stay up


Nope - I was wrong again, mysql is running (I checked as the zenoss user rather than as root :8} ) so back to
square one I guess.

--------------------------------------------------------------

To post a reply to the thread message, either reply to this email (recommended) or visit the message page: http://community.zenoss.org/message/50223#50223

--end--
Permalink | Reply | 0 comments |
headers
hackman238 | 1 Jul 2010 15:35

ZenPerfSNMP Query Blackouts


I've been finding that ZenPerfSNMP stops seding SNMP queries for 2 to 3 polling cycles at about the same time
(2:15AM) every morning. This particular system is rebooted every Tuesday at 5 AM, and has an average load
of 2.5. There are no obvious load issues. Has anyone else expereiced this?
 
Zenoss 2.5.2
Redhat Enterprise 5.4 X86-64

--------------------------------------------------------------

To post a reply to the thread message, either reply to this email (recommended) or visit the message page: http://community.zenoss.org/message/50224#50224

--end--
Permalink | Reply | 6 comments |
headers
rgartley | 1 Jul 2010 16:50

Re: zInterfaceMapIgnoreNames Problem


I know this post is a bit old but it's the only one I've found that's somewhat relevant to my issue.
 
When monitoring my PRI's on Cisco Voice Gateways, I don't want to see each and every B channel on the Serial, I
just want to see the D channel so I can monitor up/down from the telco. Zenoss is nice enough that it
recognizes this as 'Serial0/1/0:23-Signaling' on the interface description.
 
I'm using the 'standard issue' regex for filtering out interfaces and I've added Serial to it so it looks a
bit like this....
 
^(Serial|Voice Over IP Peer|unrouted|rif0|lo|vLAN)
 
As you can see, I'm ignoring everything Serial, BUT I'd like to ignore everything Serial and allow any
Serial with the word Signaling in it.
 
A few of you might ask why I want to ignore everything Serial on my routers. Well, I have a seperate device
class for my voice gateways and my voice gateways are all Ethernet and the except for the Serial PRI's.
Applying this regex on the voice gateway is no issue as the rest of my routers are in a different device class.
 
I'm not a regex guy at all so I'm looking for a little help here in nailing out the regex syntax. Any help on this
is apprecieated!

--------------------------------------------------------------

To post a reply to the thread message, either reply to this email (recommended) or visit the message page: http://community.zenoss.org/message/50228#50228

--end--
Permalink | Reply | 0 comments |

Gmane