If it's not showing these, then it may be that the devices doesn't list these with the RFC standards as it does for the interfaces. IF you use the snmpwalk.exe in the Orion install directory you can get a list of all supported OIDs. You can also use the PollerCheckerTool.exe and see if it does respond to the CPU and memory pollers. If not, the SNMP walk is your best option. You can then take these values and enter the OIDs into the Universal Device Poller to have the data polled and collected.
I poll our NetScalers for Mem/CPU usage in the form of UnDPs via the below OID's.
Hope this helps.
is their a undp that I can download or can you post yours? thanks.
I found it thanks. How do I find out what other UNDPs are possible for this or any other device?
As mharvey stated you can walk the devices in question and just do some research online to see what is available on the OID trees for the devices. And of course just check to see what is uploaded to thwack already.
I ran the snmp walk on my netscaler. how do I decide which value is the oid that can be used?
You can search online using the OIDs to see if you can find the MIB information that ties to it as far as a description, or you can use the Universal Poller and add in the OID, and if the MIB is in the Orion MIB Database, it will pull up the description of what statistics the MIB is gathering.
now I a little confused. I was told i can run an snmp walk to see which OIDs are supported. I ran the walk but I don't see the OID number. Is there a good way to look through the MIB tree and see everything relating to the netscaler. Are you also saying there may be OIDs outside of the list in the MIB tree?
There may be MIBs that are not included in the Orion MIB database yes. The MIB database doesn't have all the MIBs a vendor may possibly create for SNMP polling. It does have a large amount, but there may be some out there that are new, that haven't been presented to Orion for addition by a customer. The OID number should be in the SNMP walk text that is saved after the SNMPwalk completes.
I ran an snmp walk on my netscaler and then searched for the OID above 126.96.36.199.4.1.59188.8.131.52.41.1 and this exact string is not in the text file that was generated.
I do see other lines that are close, such as, .184.108.40.206.220.127.116.11.0 = OID: 18.104.22.168.4.1.5951.1
is this the OID number I would put into the UNDP or do I pick other variables after it that are found in the walk?
when I run the snmp walk tool it asks for the root OID and the default is 1. does changing that make a difference and how do i know which root OID I should use? thanks for your help on this, this is very good info.
You can use 1. and it will scan the whole MIB tree for device support. However, doing this can lead to timeouts depending on the amount of OIDs supported on the device, so you may want to increase the timeout settings. Another option would be setting the value to the OID you want to find such as the one mentioned by Michael Melton. From the Universal Device Poller, you'd input the OID which would be the 22.214.171.124.4.1.5951.1, not the output.
I want to be able to monitor for everything. How would I get a complete list for this device or any other device that I'm monitoring? I don't know what OIDS I want so I would like to get them all. thank you.
where are the timeout settings for the snmp walk? I can see the requests set to 20. is that the timeout?
To get them all, the SNMP walk is what you are going to want, along with perhaps discussion with device vendors. I would not recommend "everything" as there is going to be information given by SNMP that may not have any relevance to what you need to monitor. The timeout setting is on the top in the right of the snmpwalk.exe that says SNMP Timeout [ms]. By default its 2500, you may want to double or even triple that value to give the device more time to report supported OIDs. An SNMP walk is going to be the only real way to get a complete list of all supported OIDs for any device. SolarWinds, by default, only goes after industry standards for monitoring, such as Interface bandwidth, errors, CPU, and memory, using for the most part RFC1213 standard MIBs, with some exceptions.
Here is a screenshot of the snmpwal.exe console. I do not see the timeout.
How do I browse the MIB tree for everything related to a specific device such as netscaler?
can you tell me what identifies the OID number in this string from my snmp walk:
what identifies th.126.96.36.199.188.8.131.52.0 = STRING: "NetScaler NS9.3: Build 54.4.nc, Date: Dec 20 2011, 22:44:41 "
.184.108.40.206.220.127.116.11.0 = OID: 18.104.22.168.4.1.5951.1
.22.214.171.124.126.96.36.199.0 = TIMETICKS: 1757030121
.188.8.131.52.184.108.40.206.0 = STRING: "WebMaster (default)"
.220.127.116.11.18.104.22.168.0 = STRING: "NetScaler"
.22.214.171.124.126.96.36.199.0 = STRING: "POP (default)"
.188.8.131.52.184.108.40.206.0 = INTEGER: 72
.220.127.116.11.18.104.22.168.0 = INTEGER: 7
.22.214.171.124.126.96.36.199.1.1.1 = INTEGER: 1
.188.8.131.52.184.108.40.206.1.1.2 = INTEGER: 2
.220.127.116.11.18.104.22.168.1.1.3 = INTEGER: 3
.22.214.171.124.126.96.36.199.1.2.1 = STRING: "0/1"
.188.8.131.52.184.108.40.206.1.2.2 = STRING: "1/1"
.220.127.116.11.18.104.22.168.1.2.3 = STRING: "LO/1"
.22.214.171.124.126.96.36.199.1.3.1 = INTEGER: 6
.188.8.131.52.184.108.40.206.1.3.2 = INTEGER: 6
.220.127.116.11.18.104.22.168.1.3.3 = INTEGER: 6
.22.214.171.124.126.96.36.199.1.4.1 = INTEGER: 1500
.188.8.131.52.184.108.40.206.1.4.2 = INTEGER: 1500
.220.127.116.11.18.104.22.168.1.4.3 = INTEGER: 1500
.22.214.171.124.126.96.36.199.1.5.1 = GAUGE32: 1000000000
.188.8.131.52.184.108.40.206.1.5.2 = GAUGE32: 1000000000
.220.127.116.11.18.104.22.168.1.5.3 = GAUGE32: 0
.22.214.171.124.126.96.36.199.1.6.1 = HEX-STRING: 005056920043
.188.8.131.52.184.108.40.206.1.6.2 = HEX-STRING: 005056920044
.220.127.116.11.18.104.22.168.1.6.3 = HEX-STRING: 005056920043
.22.214.171.124.126.96.36.199.1.7.1 = INTEGER: 1
.188.8.131.52.184.108.40.206.1.7.2 = INTEGER: 1
.220.127.116.11.18.104.22.168.1.7.3 = INTEGER: 1
.22.214.171.124.126.96.36.199.1.8.1 = INTEGER: 1
.188.8.131.52.184.108.40.206.1.8.2 = INTEGER: 1
.220.127.116.11.18.104.22.168.1.8.3 = INTEGER: 1
.22.214.171.124.126.96.36.199.1.9.1 = TIMETICKS: 625400
.188.8.131.52.184.108.40.206.1.9.2 = TIMETICKS: 625400
.220.127.116.11.18.104.22.168.1.9.3 = TIMETICKS: 625400
.22.214.171.124.126.96.36.199.1.10.1 = COUNTER32: 0
.188.8.131.52.184.108.40.206.1.10.2 = COUNTER32: 0
.220.127.116.11.18.104.22.168.1.10.3 = COUNTER32: 922263493
.22.214.171.124.126.96.36.199.1.11.1 = COUNTER32: 38318020
.188.8.131.52.184.108.40.206.1.11.2 = COUNTER32: 0
.220.127.116.11.18.104.22.168.1.11.3 = COUNTER32: 3041703933
.22.214.171.124.126.96.36.199.1.12.1 = COUNTER32: 3340825470
.188.8.131.52.184.108.40.206.1.12.2 = COUNTER32: 0
.220.127.116.11.18.104.22.168.1.12.3 = COUNTER32: 0
.22.214.171.124.126.96.36.199.1.13.1 = COUNTER32: 0
.188.8.131.52.184.108.40.206.1.13.2 = COUNTER32: 0
.220.127.116.11.18.104.22.168.1.13.3 = COUNTER32: 0
.22.214.171.124.126.96.36.199.1.14.1 = COUNTER32: 0
.188.8.131.52.184.108.40.206.1.14.2 = COUNTER32: 0
.220.127.116.11.18.104.22.168.1.14.3 = COUNTER32: 0
.22.214.171.124.126.96.36.199.1.15.1 = COUNTER32: 0
.188.8.131.52.184.108.40.206.1.15.2 = COUNTER32: 0
.220.127.116.11.18.104.22.168.1.15.3 = COUNTER32: 0
.22.214.171.124.126.96.36.199.1.16.1 = COUNTER32: 0
.188.8.131.52.184.108.40.206.1.16.2 = COUNTER32: 0
.220.127.116.11.18.104.22.168.1.16.3 = COUNTER32: 4247892584
.22.214.171.124.126.96.36.199.1.17.1 = COUNTER32: 38447591
.188.8.131.52.184.108.40.206.1.17.2 = COUNTER32: 59732
.220.127.116.11.18.104.22.168.1.17.3 = COUNTER32: 3078314995
.22.214.171.124.126.96.36.199.1.18.1 = COUNTER32: 0
.188.8.131.52.184.108.40.206.1.18.2 = COUNTER32: 0
.220.127.116.11.18.104.22.168.1.18.3 = COUNTER32: 0
.22.214.171.124.126.96.36.199.1.19.1 = COUNTER32: 0
The timeout might not have been added until the recent build of 10.3. In regards to what identifies the values, it would be the response given back by the SNMP daemon on the device. The OIDs are sent by the SNMP walk to the device, and the SNMP daemon responds back with answers. To browse for vendor specific MIBs, such as netscaler, you'd need to know the OIDs. That you would have to get from the device manufacturer, such as Citrix, or via a web search.
OK, so with the output from the snmp walk I posted earlier how do I identify the OIDS that I can put into the UNDP to monitor something?
You can put any of them in there such as 188.8.131.52.184.108.40.206.1.16.3. If there is specific information you are looking for, you'd need to web search for the specific OIDs, or just add in the OID to the Universal poller to see what the MIB is and a description of it.
when I put in that number I don't get a description does that mean it's just not in the MIB database? Is there a way to see a description or do I have to add it and see what it says on the solarwinds webpage?
Looking at that number in the output I can see that it says = counter32. is the counter32 what I look for or does that matter?
If there is no description when you put in the OID then it means the MIB is not in the SolarWinds MIB Database, if the device still answers to it, that is ok, it doesn't have to be in the database. You'd need to research the OID to deteremine what the value is. The counter32 is the type of data that the MIB represents. In this case it is a 32 bit counter, so the value will reported is 32 bit data and is a counter that should continuously increase until it rolls over. The data in a counter that is relevant is the delta, or value between poll A and B. When creating the poller you can go into the Advanced Options and set the data value as a counter so that the statistic reported is that delta.
Thanks! Great info! I have (2) Netscaler appliances that just came in a month or so ago.
Much appreciate it!