Dell server reboots when auto self-test is initiated

[Anonymous user]

My Dell T310 server is powered through a BR1000G UPS. The server reboots itself every 14 days exactly. I have proven through troubleshooting that the UPS is the culprit. Since this is the interval for the auto self-test, what is going on? Is there a fix for this?

S/N: 3B1241X00207

Firmware version: 868.L3.D

[BillP]

Hello,

We typically recommend server class hardware to utilize a pure sinewave output UPS. Power supplies in servers often cannot deal with the stepped approximation to a sinewave a Back-UPS outputs on battery and/or cannot deal with the transfer time to battery. Dell may also have a recommendation on what type of sinewave input is required for this server power supply.

Out of curiosity, does it do it when you put the UPS on battery manually (i.e. pull the plug from the wall) or only during the auto self test? Did this just develop over time?

Further detail we have on this topic is here -> http://www.apc.com/support/index?page=content&country=ITB〈=en&locale=en_US&id=FA158811 and I'd suggest reviewing "problem" #5 and #6 (if this used to work properly). Based on what you said so far, I don't think this is a "problem" that you need to fix but a possible incompatibility with the output sinewave type of the UPS and what the Dell power supply accepts. I went here -> http://www.dell.com/us/business/p/poweredge-t310/pd and browsed for a UPS for this server and it recommended to me all Smart-UPS products. You may want to consider checking with Dell on this server's power requirements too before you do anything to do see what they say.

[Anonymous user]

Angela, thanks for the prompt reply. I have referred to my notes to try to reconstruct the history of this issue. This Back-UPS was installed in December, 2012. There were no indications of anything unusual until 6 months later (June, 2013), when periodic server restarts began. At that time, I did not discover any repeating cycle time or cause, and the problem eventually went away. It wasn't until December, 2014 that this issue reappeared. I began working with Dell support last month to target any potential server issues and none were found. I independently discovered that the server dropouts were caused by the self-test interval of the UPS. Dell's response was to suggest that I see if there was a firmware update to the UPS. So there is no widely held knowledge at Dell that there may be a problem with particular models of your products. I will, however, forward your comments to them for their response to see if they will confirm a product mismatch. I have not tried to manually test switchover to battery power, but may do that for information sake.

[BillP]

Hello Leslie,

No problem. And if you want, tell them I used the Dell UPS selector on www.dellups.com (which I think is a website that we make for them based on our UPS selector). 

Let me know if I can help further or provide any additional information. (These Back-UPS don't support field firmware updates by the way even if that was what needed to do.)

[Anonymous user]

Hello, Angela,

I was finally able to perform a manual switchover from AC to battery power by "pulling the plug" on the UPS. The UPS switched to battery power and the server continued to operate with no interruption. After operating OK for 7 minutes, I switched back to AC power. This test proves to me that the server has no problem running on squared sine wave input power and that the switchover time is not an issue. The problem lies specifically in the self tests that are performed every 14 days. Those self tests are causing the server to momentarily lose power. This is obviously a failure mode for the Back-UPS Pro product and I am disappointed that APC does not recognize it as such. This unit recently passed its 2-year warranty date, and will be replaced by a non-APC product. 

[BillP · Accepted solution]

Hi Leslie,

I understand your frustration with this. It does not make too much sense to me and not something we see regularly. If it was, it'd be surely addressed. And it also seemed to develop over time from what you mentioned.

I know it sounds like you don't want to deal with it further but it could also be something related to the USB connection maybe but doubtful based on the pull plug test. Just a thought. All a self test does is an automated transfer to battery to compare the voltage drop on the battery versus the UPS load to see if it remains within a certain tolerance range. 

Please let us know if you wish to discuss this further. If it was a failure in the UPS that was an inherent problem, we'd certainly attempt to address it but it may require further detail on the scenario and your observations. This is not something customers see and at this time, I'd say it is an isolated case we'd work with you on to see if we can pinpoint the cause.