Multiple Questions on the Netbotz 250

DCIM_Support · ‎2020-07-04

New to the Netbotz 250.

Had a 355 fail and testing a 250 as the replacement for it (the 355 has features never needed/used such as camera, air flow, dew point).

Standalone unit, using the web interface, no dco/dce to go along with it.

Several questions... (of these, #2 is possibly the biggest concern)

1) Bold warning about plugging up to a PoE switch. I think we're hard pressed to find any enterprise level switch that doesn't do PoE. So, realistically, will it harm the NB250 to plug it into a switch supporting POE? (I don't have switch access so cannot disable PoE). Right now as I test it, it's plugged into a small deskside switch.

2) When I look at the data logs for temperature (wired sensor that came with the unit), I see temp readings that are about 7 degrees less than what the sensor shows when I look at the status. It's been online for about 72 hours as of now and this discrepancy looks to be consistent. I see the same thing when I graph the data log or export it to a csv file. For example, it shows 66.4 degrees when I look at live status. But the logs all show temps of less than 60 degrees. Interval is set to 10 minutes. If there was wide variation in temp then I could see the average being less than the current, but that is not the case.

Running the latest firmware - 6.5.6

This was a test... and I might just go back to the 355 as there were some nice features w/o having to get other software (for example I could have it email me the graphics on a scheduled basis).

Thoughts?

Attached a couple images. One of the 8 hr graph. One of the current status of the same sensor.

Humidity is logging and showing correctly, so it's only the temp that is logging/graphing wrong.

And yes, I know that the temperature is NOT hovering around 57 😀

I'll connect a couple more on Monday to see if this is consistent.

DCIM_Support · ‎2020-07-04

I added 3 more wired temp/humidity sensors this morning.

All of them show the same symptom described in #2 above. The "live" temperature reading does NOT match the readings being logged/graphed. (I looked at the graph and I exported the data and looked at it in the CSV file). These same sensors worked and logged properly on the old 355 that I had.

Here is another sample from a different sensor this morning.

DCIM_Support · ‎2020-07-04

Dear Cal H,

Very interesting problem, similar to a bug: please take temperature/humidity readings in NetBotz-250 every minute and compare with the "live" temperature/humidity of the wired sensors. What will be the result?

Please specify the model of your wired temperature/humidity sensors?

With respect.

DCIM_Support · ‎2020-07-04

Thanks for that suggestion. I am doing that now. Based on maybe 10 minutes at the 1 minute interval, the logged temp (not so much the humidity, just the temp) is still skewed. Image attached of a 3 minute window, and of the live data as of the most recent 1 minute log.

Interestingly, I added a wireless temp sensor (the one that came with the package). It's temp is logged correctly at the 10 minute interval when I compare the logged data to live data. It's also matching at the 1 minute interval.

The wired temp/humidity sensors are the AP9335TH ones - 3 'older' that I already had, 1 new that came with the 250.

I opened a support call, and spent an hour or so in a WebEx. They can see what I'm seeing, but when they tried to reproduce on their 250 with the same FW version they don't seem to be able to reproduce...

DCIM_Support · ‎2020-07-04

Dear Cal H,

Ok, then another experiment - change the unit of measurement from F to degrees C. If you get a similar bug, then most likely it is the hardware problem of your individual NetBotz-250...

With respect.

DCIM_Support · ‎2020-07-04

Another good suggestion. I changed from F to C. Watched a few minutes of the logging and it appears to be the same. So I rebooted the 250 and then watched a few more minutes of logging after a restart just to be sure.

I'll add these notes to the existing case.

I do plan on spinning up a DCE to see if it shows anything different when pulling information from the 250. I suspect SNMP queries will pull live data, so that might not tell me much.

Will be interesting to see how this is handled as right now it makes my logged data/history useless.

DCIM_Support · ‎2020-07-04

Dear Cal H,

I do plan on spinning up a DCE to see if it shows anything different when pulling information from the 250. I suspect SNMP queries will pull live data, so that might not tell me much.

Will be interesting to see how this is handled as right now it makes my logged data/history useless.

Ok, find out how it will be in DCE software, you can already now and quickly 😀. To do this, simply download the snmpwalk-package from the web-page How to create a SNMP walk file for Struxureware Expert/InfraStruXure Central. And follow the instructions to create a snmpwalk-file for your NetBotz-250 with your firmware.

After analyzing the snmpwalk-file (this is simple ASCII file), you can find out exactly which OIDs are corresponding to your wired temperature/humidity sensors (not wireless, as they are detected separately). I can help you with this and you will be able to find out whether SNMP values correspond to real values or not.

With respect.

DCIM_Support · ‎2020-07-04

Done. SNMPwalk output created. I'm looking through it and also looking for a reference to see which OIDs are the actual temp/humidy values.

DCIM_Support · ‎2020-07-04

I.e., even through SNMP you see a mismatch with the real temperature?

DCIM_Support · ‎2020-07-04

After a little digging, I'm guessing celcius, and these relevant values based on where my sensors are:

1.3.6.1.4.1.318.1.1.10.4.2.3.1.5.0.1 (temp in celcius rounded sensor port 1)

and ........6.0.1 (humidity rounded sensor port 1)

If that's the case then SNMP is pulling the current values rounded to the nearest whole number

DCIM_Support · ‎2020-07-04

Rounding must be up to 0.1. Is this true for your NetBotz-250?

Also, does OID tree 1.3.6.1.4.1.318.1.1.10.4.2.3.1.9 exist in your snmpwalk-file?

With respect.

DCIM_Support · ‎2020-07-04

Yes, that OID exists... and it looks like the temp in celcius w/o a decimal – so 181 for example (matching 18, and 165 matching 16.5 and rounding to 17).

I assume humidity is just rounded to nearest whole #

Looks like wireless sensor data is probably under 1.3.6.1.4.1.318.1.1.10.5.1.1.1 or similar.

Thanks!

DCIM_Support · ‎2020-07-04

Dear Cal H,

This is correct, because the NetBotz-250 via SNMP protocol sent the temperature of wired sensors twice: with normal accuracy and with high precision. Those, 181 - 18.1C, 165 - 16.5C, etc. In DCE software, these values will be displayed with an accuracy of 0.1.

With respect.

DCIM_Support · ‎2020-07-04

Re #1, I answered that one myself. Plug it into a POE port and it never comes online. You get link, and activity but it never comes online network-wise.

Bad design (in my opinion) on that, as in enterprise networks, most enterprise switches are POE now in order to support VoIP and often cannot be managed by local teams (as is my case). So my only option is to either get the management team to disable POE on the port, or put a non-poe switch in between the enterprise port and the 250.

DCIM_Support · ‎2020-07-04

there are different levels of POE - if you need "AT" and it isn't being offered... things won't work

802.3af - 15.4 watts maximum

802.3at - 30 watts maximum

802.3bt - 60 watts maximum (this requires 4 pairs of wire)

regular POE is 2 par, strands 4,5 and 7,8 (the blue and brown pairs)

DCIM_Support · ‎2020-07-04

Dear Cal H and Eric Clark,

I agree with Cal H and very much hope, that this bug will be fixed in future revisions of NetBotz-250 and its firmware. But since this can be a very lengthy process, I can only recommend using two-pair patch cord (only green and orange pairs) to connect the NetBotz-250 via ethernet.

For example, it could be a standard Cat. 5e patch cord with blue and brown pairs cut off from one side. Or you can use the standard RJ-45 coupler (F-to-F type) with 4,5 and 7,8 contacts removed. All this will allow, without costs, to avoid the above mentioned problem in NetBotz-250 right now.

Dear Cal H, we will be interested to know if you could solve your problem with wire temperature/humidity sensors or not?

With respect.

DCIM_Support · ‎2020-07-04

In this case, the 250 simply doesn't work with a PoE port (it really should - and should simply ignore the probe as the switch will not send power if not 'requested'. Its a shame that the workaround is to have a modified 2 pair cable so as not to have the issue... or to have to inject a non-poe switch between an enterprise one and the 250.

Regarding the wired temp/humidity data logging, no solution was identified. We can confirm that the wireless sensor reads correctly and that the data is logged correctly. Fir the wireless, temp/humidity, the sensor reads correctly 'real-time'. SNMP queries pull the correct data 'real-time'. But the logging of the temperature is many degrees off - but not the humidity - it is correct. Logging at 1 minute intervals shows the same variation - 6-10 degress lower than the actual temperature is what is logged.

I have updated the ticket several times but no further response, so today I will likely call back directly.

Interesting fact about my failed 355. I took it apart to see if I could identify a blown part/etc. In the middle of the board is a typical device power connector - it's even labeled AUX PWR. And lo and behold, the plastic back even has a hole cut in the plastic to accommodate it, but the adhesive label covers it up. I went so far as to cut the label and locate a suitable AC adapter, and the 355 still works perfectly fine when powered via the AUX PWR. It just will not work when you try to power it via POE now. I cant find that AUX PWR documented anywhere and sure could not find an adapter for sale – just thought I'd mention it. I'm considering putting it back into use if nothing is resolved on the 250.

DCIM_Support · ‎2020-07-04

From your comment:

In this case, the 250 simply doesn't work with a PoE port (it really should - and should simply ignore the probe as the switch will not send power if not 'requested'.

I doubt it. It should not be. Have you checked this? Which model of PoE switch do you use?

...But the logging of the temperature is many degrees off - but not the humidity - it is correct. Logging at 1 minute intervals shows the same variation - 6-10 degress lower than the actual temperature is what is logged.

I have updated the ticket several times but no further response, so today I will likely call back directly.

This is a really problem, and Schneider Electric employees must answer: either this is a problem with the latest firmware or it is a malfunction of your NetBotz-250, which you just need to replace under warranty.

...I went so far as to cut the label and locate a suitable AC adapter, and the 355 still works perfectly fine when powered via the AUX PWR. It just will not work when you try to power it via POE now.

Many thanks for the interesting and useful information - I did not know about it. This is a really good solution to the problem. What voltage and amperage did you use AC adapter?

With respect.

DCIM_Support · ‎2020-07-04

I'm using a 13V, .8A adapter. It also worked at 12V, .3A when I initially tested the connector. Given that it is not documented anywhere that I have been able to find, I'm not sure what the correct / recommended V / A is.

DCIM_Support · ‎2020-07-04

Ok, many thanks for the information 😀.

I will also be interested in how your problem will be solved with wire temperature/humidity sensors and the NetBotz-250 connection to the PoE switch.

With respect.

DCIM_Support · ‎2020-07-04

FYI... I heard back... the problem with the logging of different values for temperature on wired Temp/Humidity sensors (different than the actual temperature) was identified and should be fixed in a new firmware release in August-September. No indication of what the actual problem was, but it is being addressed.

DCIM_Support · ‎2020-07-04

Dear Cal H,

Thanks for the useful information. I understand your problem, so let's hope that soon this bug will be fixed.

With respect.

DCIM_Support · ‎2020-07-04

Dear Cal H,

In connection with the release of the new firmware v.6.6.7 for the NetBotz-250 device, I can say the following about your questions.

Q1: As stated on the official resource of User Assistance for NetBotz Rack Monitor 250:

- IMPORTANT! The NetBotz Rack Monitor 250 is not a PoE compatible device. Do not connect a NetBotz Rack Monitor 250 to a PoE (Power over Ethernet) switch.

Q2: As stated in the release of the new firmware v.6.6.7 Release Notes.pdf:

That is, in the new firmware version v.6.6.7 for NetBotz-250 devices the above specified bug is fixed.

Try it and report the result, please.

With respect.

DCIM_Support · ‎2020-07-04

Hi Cal H,

Just an FYI on the POE thing, I do not see in the report on that issue that there was any damage specifically reported however there were a number of reports where the device would simply not come up on the network. I would simply avoid it if possible as noted in previous responses.

Steve

DCIM_Support · ‎2020-07-04

This question is closed for comments. You're welcome to start a new topic if you have further comments on this issue.