Send OID description:
$ snmptranslate -Td ENTITY-SENSOR-MIB::entPhySensorScale ENTITY-SENSOR-MIB::entPhySensorScale entPhySensorScale OBJECT-TYPE -- FROM ENTITY-SENSOR-MIB -- TEXTUAL CONVENTION EntitySensorDataScale SYNTAX INTEGER {yocto(1), zepto(2), atto(3), femto(4), pico(5), nano(6), micro(7), milli(8), units(9), kilo(10), mega(11), giga(12), tera(13), exa(14), peta(15), zetta(16), yotta(17)} MAX-ACCESS read-only STATUS current DESCRIPTION "The exponent to apply to values returned by the associated entPhySensorValue object.
*This object SHOULD be set by the agent *during entry creation, and the value SHOULD NOT change during operation."
On Fri, Jul 3, 2015 at 4:20 PM, Robert Williams Robert@custodiandc.com wrote:
Yay J
Thanks for that, for what it’s worth, the unit allows up to 9 sensors in the chain I believe (18 inputs if they are all Temp+RH combined), so whatever you have fixed for the .1 and .2 OIDs it should be set to apply up to .18 for anyone else who may have more sensors attached than us.
If you are able to let me know the OIDs where the scaling and descriptions ‘should’ be for those sensors then I’ll poke Riello about implementing it properly.
Thanks again!
Robert Williams Custodian Data Centre Email: Robert@CustodianDC.com http://www.CustodianDC.com
*From:* mike@stupalov.ru [mailto:mike@stupalov.ru] *On Behalf Of *Mike Stupalov *Sent:* 03 July 2015 14:12 *To:* Robert Williams *Cc:* Observium Network Observation System
*Subject:* Re: [Observium] Scaling adjustment for UPS sensors
Done, try in r6555. :)
On Fri, Jul 3, 2015 at 4:06 PM, Robert Williams Robert@custodiandc.com wrote:
Hi – thanks for looking, yes there is a CLI and a Web portal, which looks like this:
It does seem to be a rather rubbish implementation, including the fact that the sensors keep mis-reading due to EMF interference (ironically, coming from the UPS itself!). Riello are working with us to resolve that so while I have their attention I’ll see if we can get the SNMP implementation improved!
There is nowhere in the CLI to ‘name’ the sensor either, they are just numbered as per the image above, which is only based on the order they are detected in… nice…!
Cheers :)
Robert Williams Custodian Data Centre Email: Robert@CustodianDC.com http://www.CustodianDC.com
*From:* mike@stupalov.ru [mailto:mike@stupalov.ru] *On Behalf Of *Mike Stupalov *Sent:* 03 July 2015 13:55 *To:* Robert Williams; Observium Network Observation System
*Subject:* Re: [Observium] Scaling adjustment for UPS sensors
You have any webinterface (or cli) on UPS which shows the temperature and humidity from device?
I set scale 0.1 also for humidity, because not sure that 60% RH correct :)
Also note, sensor name set as "Chassis", because device also not report correct Name for it:
CMD[/usr/bin/snmpwalk -t '5' -r '3' -v1 -c *** -Pu -OQUs -m ENTITY-MIB:CISCO-ENTITY-VENDORTYPE-OID-MIB -M /opt/observium/mibs/rfc:/opt/observium/mibs/net-snmp:/opt/observium/mibs/cisco 'udp':'my-de vice.local':'161' entPhysicalDescr]
EXITCODE[1] RUNTIME[0.0446s] STDOUT[ mib-2.1.1.0 = "NetMan 204" ] STDERR[ Error: OID not increasing: entPhysicalDescr
= mib-2.1.1.0
] SNMP_STATUS[FALSE]
On Fri, Jul 3, 2015 at 3:18 PM, Robert Williams Robert@custodiandc.com wrote:
Hi Mike,
Thanks for that! Good news - It’s now showing temperature correctly, bad news – it has dropped the RH from 60% to 6% which is now incorrect. I will feed the lack of scaling information back to Riello and see if they can sort it (we are on a roll with them at the moment).
FYI to help clarify the statistics displayed: The ‘Battery’ temperature which the UPS outputs is actually *always* the Chassis itself. The Riello units never give out their battery temperature over SNMP, it’s always the unit’s supply air for itself (even when you fit an additional battery charge circuit temperature sensor to it, like we have, it still gives you chassis air temperature over SNMP and not the batteries).
Also, the NetMan sensors we have fitted are just ‘loose’ so can be placed anywhere, in our case they are next to batteries but for other people this may vary, so I wouldn’t call them anything specific like Chassis or Battery. You may be better off labelling the external sensors as something generic (like External Temp 1, 2 etc.) perhaps?
Just my 2p worth J
Cheers!
Robert Williams Custodian Data Centre Email: Robert@CustodianDC.com http://www.CustodianDC.com
*From:* observium [mailto:observium-bounces@observium.org] *On Behalf Of *Mike Stupalov *Sent:* 03 July 2015 12:54
*To:* Observium Network Observation System *Subject:* Re: [Observium] Scaling adjustment for UPS sensors
Ok, I looked the output and made fixes in r6554.
The problem on your device is that it not report sensor scales by ENTITY-SENSOR-MIB, which should not be.
Try latest revision, of course after 'svn up' run rediscover:
./discovery.php -h <your_device>
(also added some new sensors by UPS-MIB)
On Fri, Jul 3, 2015 at 10:35 AM, Robert Williams Robert@custodiandc.com wrote:
Hi,
A quick update on this, I’ve sent the debug to Mike, to save spamming the list (let me know if anyone else wants it!). The RH is correct as the sensor is in the battery room (not on the data floor) and rather ironically I took the screenshot during the hottest day of the year where we are and that room is chilled to 20C so RH goes up :)
Cheers,
Robert Williams Custodian Data Centre Email: Robert@CustodianDC.com http://www.CustodianDC.com
*From:* observium [mailto:observium-bounces@observium.org] *On Behalf Of *Tom Laermans *Sent:* 02 July 2015 21:16 *To:* Observium Network Observation System *Subject:* Re: [Observium] Scaling adjustment for UPS sensors
7.3% rH is way too low though.
Riello is notoriously shit in their SNMP implementation though (I have them too).
I see it's a Netman 204, that should be quite decent, but afaik those sensors are using regular IEEE UPS-MIB.
I don't have any attached to mine (not sure if I even can) - if anyone wants to send me a sensor I'll gladly attach it to one of my Riello UPSes ;>
Tom
On 07/02/2015 05:28 PM, Mike Stupalov wrote:
On 02.07.15 18:25, Phillip Baker wrote:
Are you sure your relative humidity is fine? 73% is.. well, pretty bloody wet. Is the humidity probe in a shower room? Anything over 60% should definitely be considered abnormal in most datacentre environments.
Here same problem with scalling.. I think this is 7.3%
.. still waiting for debug output
*From:* observium [mailto:observium-bounces@observium.org observium-bounces@observium.org] *On Behalf Of *Robert Williams *Sent:* 02 July 2015 14:26 *To:* Observium Network Observation System (observium@observium.org) *Subject:* [Observium] Scaling adjustment for UPS sensors
Hi,
Is there a way to adjust the re-scaling for auxiliary UPS sensors? We’ve just been given some external sensors for our Riello units to play with; and whilst Observium can indeed see them, their temperatures are out by one decimal place:
(it should be 19.5C, obviously)
The RH is fine interestingly, and this is only affecting the ‘external’ sensors in the NetMan card, not the built-in sensor. Interestingly, Riello also give out the ‘chassis’ temperature under the ‘battery’ temperature OID, which is why it shows our batteries as 27 degrees, they are not - and this is not Observium’s fault J
Anyway, is there a manual ‘tweak’ to correct the scaling for the temperature?
Cheers!
Robert Williams Custodian Data Centre Email: Robert@CustodianDC.com http://www.CustodianDC.com
observium mailing list
observium@observium.org
http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
--
Mike Stupalov
observium mailing list
observium@observium.org
http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
--
Mike Stupalov http://observium.org/
--
*Mike Stupalov http://observium.org/http://observium.org/ http://observium.org/*
--
Mike Stupalov http://observium.org/
