-
Re: Report to find out servers not rebooted for 'n' number of days
bkyle Oct 11, 2017 12:49 PM (in response to dhyanesh.mehta)Dhyanesh,
If these devices are using SNMP, you could try a report based on system up time (1.3.6.1.2.1.1.3).
-
Re: Report to find out servers not rebooted for 'n' number of days
dhyanesh.mehta Oct 16, 2017 4:14 AM (in response to bkyle)Hello Kyle,
Thanks for the reply. Yes all devices i want to monitor has SNMP. I already tried to add Uptime in the report but issue i am facing is it gets all monitors in Uptime not jus Ping. For example what i did was i go to report> Add custom report> Select group of Device> i selected a group which contains multiple devices each have multiple monitors like Ping, drive space, Ram, CPU and Windows Services. Now when i select Uptime parameter for Last Month and click on preview it brings status report of all monitors which are up since last month that includes Ping, Drive Space, Service and all.
What i am interested is only Ping so i can determine this device has been up and running since last 30 days or not. So can you please guide me how can select Ping only? I could not fine Server Uptime as parameter in Report only Uptime is there.
Thanks,
Dhyanesh Mehta
-
Re: Report to find out servers not rebooted for 'n' number of days
dhyanesh.mehta Oct 30, 2017 4:51 AM (in response to bkyle)Hello Kyle,
Can you please guide me how can i generate report as per SNMP string you provided? I am not that much expert at it so if you can list down the steps it will really help.
Thanks,
Dhyanesh Mehta
-
Re: Report to find out servers not rebooted for 'n' number of days
dhyanesh.mehta Nov 6, 2017 7:33 AM (in response to bkyle)Hello BKyle,
I finally figured out what you mentioned by sysuptime SNMP. I successfully added sysuptime custom SNMP monitor.
1 question i have now is it gives result in Seconds. Is it possible to configure it to give result in Days? So when any monitor is down and we check it shows directly this device has not been rebooted since 40 days instead of x seconds.
Thanks,
Dhyanesh Mehta
-
-
Re: Report to find out servers not rebooted for 'n' number of days
dhyanesh.mehta Oct 29, 2017 2:03 AM (in response to dhyanesh.mehta)Hello,
Its been a while since i did not get any help here. Please guys somebody help me out here.
Thanks,
Dhyanesh Mehta
-
Re: Report to find out servers not rebooted for 'n' number of days
malasherv Oct 30, 2017 6:33 AM (in response to dhyanesh.mehta)Hello,
You can try this, add the sysUptime OID in UnDP and monitor it for the group of devices you want, you can further take a report of the UnDP and filter for the ones with time less than 30 days.
Not sure how far this would work, scanning system Uptime would be the best guess.
Thanks,
Malcolm.
-
Re: Report to find out servers not rebooted for 'n' number of days
dhyanesh.mehta Oct 30, 2017 8:41 AM (in response to malasherv)Hello Malcom,
Thanks for the reply. Just to clarify is UnDP configuration for NMP or IPMonitor too as i could only find information about UpDP in NPM product so far.
if it is possible to configure UnDP using SNMP OID in IpMonitor can you please tell me how or share a document?
Thanks,
Dhyanesh Mehta
-
Re: Report to find out servers not rebooted for 'n' number of days
wizard_from_oz Feb 24, 2018 7:07 PM (in response to dhyanesh.mehta)Hello,
I assume you resolved this. For anyone else interested, we achieved the equivalent by adding a "SNMP: USER EXPERIENCE" monitor to each server based on 1.3.6.1.2.1.1.3.0 with a numeric check to be less that the value for the period to be monitored - in seconds. If we wanted to report, we could create a smart group with just those monitors and report on the group.
A the monitor records the SNMP value, if we wanted to see servers up more than any specific time, we could add a filter to the smart group to check for the value returned exceeding the number of days, converted to seconds. It is not as neat as being able to specify a value in days, but it works pretty close.
-
Re: Report to find out servers not rebooted for 'n' number of days
mesverrumFeb 24, 2018 7:26 PM (in response to wizard_from_oz)
One caution to using that SNMP counter is the number rolls over after enough time passes, I believe its just a bit short of 2 years, so systems up longer than that will show as having rebooted. Servers rarely stay up that long uninterrupted but network devices and appliances are more likely to do so.
-
-
-
-
Re: Report to find out servers not rebooted for 'n' number of days
cscoengineerFeb 25, 2018 11:42 AM (in response to dhyanesh.mehta)
For windows servers, I would look at the LastBoot setting. This should also be present for all SNMP devices. If not, then I would resort to the UnDP.
Just be aware that on SNMP monitored devices, if the agent was restarted - the last boot will reflect the time when the agent was started.
The code is same as Node DOwn with duration code with changing the eventtime to lastboot. I added a condition for just windows servers; this can be changed based on need.
I've had clients set up several of these. One for Networking, one for Servers, etc.
SELECT
-- Node
n.Caption AS [Node Name]
,DetailsUrl AS [_LinkFor_Node Name]
,'/Orion/images/StatusIcons/Small-' + StatusIcon AS [_IconFor_Node Name]
-- IP and availability
,n.IP_Address as IP_Address
,'/Orion/DetachResource.aspx?ViewID=2123&ResourceID=21222&NetObject=N:'+tostring(n.nodeid) as [_LinkFor_IP_Address]
,'/Orion/Images/SubViewIcons/VitalStats03.png' as [_IconFor_IP_Address]
--edit
--,'Edit' AS [Edit]
--, '/Orion/Nodes/NodeProperties.aspx?Nodes=' + ToString(n.NodeID) AS [_LinkFor_Edit]
--,'/Orion/images/nodemgmt_art/icons/icon_edit.gif' as [_IconFor_Edit]
--,n.CustomProperties.City as [City]
,concat(SUBSTRING(tostring(n.LastBoot),1,4),SUBSTRING(tostring(n.LastBoot),5,2)
,SUBSTRING(tostring(tolocal(n.LastBoot)),12,8)) as Downtime
,CONCAT(HOURDIFF(tolocal(n.LastBoot),getdate())/24,' Day(s) ',
HOURDIFF(tolocal(n.LastBoot),getdate())-(HOURDIFF(tolocal(n.LastBoot),getdate())/24)*24,'h ',
MINUTEDIFF(tolocal(n.LastBoot),getdate())-(MINUTEDIFF(tolocal(n.LastBoot),getdate())/60)*60,'m') AS [LastBoot]
FROM Orion.Nodes n
WHERE STATUS = 1 and n.Vendor like 'Window%'
--and n.caption like '%${SEARCH_STRING}%' or n.IP_Address like '%${SEARCH_STRING}%'
ORDER BY MINUTEDIFF(tolocal(n.LastBoot),getdate()) desc
Thanks
Amit
-
Re: Report to find out servers not rebooted for 'n' number of days
wizard_from_oz Feb 26, 2018 10:18 AM (in response to dhyanesh.mehta)Lets get back on track.
1) SNMP is the widest implemented option that allows for the same monitoring to be used - which is why we chose it.
2) Windows LastBoot, from memory, is a WMI variable requiring a different monitor arrangement for Windows, the fun and joy of WMI, and and non windows servers / devices would end up with SNMP based one anyway..
3) Yes - SNMP is not perfect, it is a 32 bit counter so will rollover when it reaches the upper limit.
4) We are talking about IPMONITOR, not ORION. IPMONITOR does not monitor uptime via any OID or WMI by default.
Our solution was the simplest to be able implement across the board, and as IPMONITOR is primarily SNMP based, seemed like the easiest and most practical option.