I’m still waiting on getting an automatic transfer switch so that I can fail-over my UPS units on my network stack. In the meantime, if I’m around, I can manually power down my network to a reduced power footprint and if needed, cut-over to a high-endurance network stack. We had one such extendeed power outage today and, upon powering the network back up, I ran into a curious/alarming result: the green status LEDs on both of my NAS units were blinking…

• Setting the stage

• Power outage!

• Blinking status!

I have two UPS units — the long term plan is to ‘combine’ them into an Automatic Transfer Switch (ATS) so that I can fail over to a secondary UPS in case one fails and also for ~double the runtime in the event that there is an outage. I’m still waiting for the ATS to come in. When I previously rebuilt the network cart, I consolidated the wiring for my NAS units — both units plug into a single compact power bar and that power bar plugs into the UPS. Something like this:

In normal every operation, everything wooks the way you would expect it to:

• Both NAS units obviously receive power

• The USB connection allows the DS1817+ to get a reading of how much battery life remains and the NAS can self-power down before the battery on the UPS drains

• The DS1817+ acts as a network UPS server for the DS920+ allowing the DS920+ to also power down before we fully drain the battery on the UPS

[Back to Top]

We recently had a two-hour power outage — just as Nicole was getting started on an important call (of course). When it was obvious that this was going to be a longer-than-usual power outage, there were a few things I could do to extend my runtime:

• Power down NAS units

• Power down extra networking gear (POE equipment, extra APs, switches that aren’t being used etc.)

With the easy things powered down, my UPS was reporting about 1-hr runtime. Thankfully there was a quick break in the call that allowed me to fully transition over to Endurance-Mode. Running on the Endurance network stack, we were good for roughly 10-hours of uptime.

On the power/UPS front, I manually moved all of the power cables from UPS A to UPS B. So things look like this:

[Back to Top]

After the utility power was restored, I decided it might be a good idea to continue using UPS B to power the network stack: this way I get a bit more even ‘wear’ on the circuitry (both UPS models are identical so they can handle the load). I powered down the Endurance stack and turned the mian network stack on.

Once the network came up, I powered on both NAS units and this was when things got strange:

• I could see both NAS units get their statically assigned IPs

• I could ping both NAS units

• But I could not connect to the management portal either by hostname nor by IP

• Using the ‘find NAS’ methods got me nothin.

A Clue from Earlier

At the start of the power outage, I powered down the DS1817+ and then the DS920+. The DS1817+ powered down without a fuss, but the DS920+ seemed to not want to power down. I tried to remote into management interface, but it wouldn’t connect.

At the time, I wasn’t in network-connectivity-troubleshooting mode, so I didn’t think about it much but the symptoms were the same: I could ping the NAS IP, but for some reason, just could not connect to the portal by any means (I ended up performing a hard-shutdown of the DS920+)

Looking at the NAS units though, I noticed that the status light was blinking. On both of them.

Synology’s help page (as of November 21, 2022) tells us that STATUS: Blinking Green means one of three things:

1. No drive installed, or

2. DSM not installed, or

3. ‘Configuration lost’

All three possibilities are quite alarming but it turns out that the actual root cause in my scenario is much more comical.

The root cause was forgetting to move the USB cable over! When I properly moved the USB cable form UPS A to UPS B, both NAS units came online and everything was running smoothly as one would expect.

When I initially moved everything from UPS A to UPS B, the original plan was to move everything back after the power was restored and had I done that, none of this excitement would have happened. Even still, this is still a bit perplexing:

• The USB cable was plugged into UPS A — which was powered down

  • How could the NAS possibly know that the other end of the USB cable was a UPS?

  • I would have expected the DS1817+ to boot up, thinking that there was no UPS connected

  • Similarly, I would have expected the DS920+ (using the DS1817+ as a UPS server) to also boot up, thinking there was no UPS conneced

• In the event that neither of them boot up, I would not expect them to respond to network pings

What about that clue from earlier?

It looks like when I powered down the DS1817+ first, it took the UPS server down with it — causing the DS920+ to be in a situation where it’s expecting a UPS server connection that is no longer there, so it then doesn’t know what to do and as such, gets stuck in a limbo state? I dunno. Lesson learned: power down all of the dependant NAS units first!

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.