debugged MODL LAN

Shortly after arriving at the Carr Astronomical Observatory, Dietmar informed me that his internet connection was not working at his My Own Dome Lot.

He had arrived the CAO Thursday but with the poor conditions had not done any work in his POD. Now, with Friday evening looking promising, he was getting things ready. It was at this point he discovered something wrong. Lightning strike? Loose connector? Flooded conduit (already)?

Dietmar told me he was using a newly purchased cable. He asked if CAT 6 was OK. Sure. I asked him if it was a straight-thru or cross-over. He wasn't sure. Even I didn't know if it would matter: probably his computer and/or the new switch would auto-sense and flip. He also reported that he was seeing a network status indicator of 100 Mbs speed. Oh really? That suggested a partial connection, likely his computer to the MODL communication junction box. A good sign. So, the problem was probably upstream from this point.

I bought the ASUS netbook to his dome, turned off the wireless, and tested with his new cable. Nothing. He fetched another ethernet cable, a known good, but it didn't work either. Again, I suspected the new MODL switch, inside the junction box, was OK, so suggested we head to the next key point, upstream, the equipment in the Geoff Brown Observatory. Besides, I didn't feel like opening the cover...

Dietmar grabbed his key set while I looked up my supervisor GBO security code. As we walked into the Warm Room, I instinctively hit the battery light switch. Didn't need it, so I powered it off. Dietmar started nudging the wood jam under the door, before activating the main power to the building. Struck me as a little odd... And then it dawned on me... Light bulb! (Sorry!) And I looked at the 3Com switch under the counter to confirm my brain wave. Yep. No LEDs.

"Dietmar," I said, "I know why your internet connection is not working...," pointing to the 3Com switch. "No power."

We threw the main lever and watched the switch power up. The nodes lit up: the input from the house, the Dynex switch for the MODLs. And a third! Right. The (new) Linksys wireless router.

I pointed out to Dietmar that the problem was further exacerbated by the wireless access point being unpowered as well. That in his POD was furthest from the house, he was struggling with a single, distant wireless signal.

I thought it very interesting, now that Dietmar has his own POD, that he doesn't need the GBO now, and had not even unlocked it.

Funny how none of us, Tony, Charles, Phil, Dietmar, myself, realised any of this. Charles had nailed it for the power at the MODL junction box. But we all forgot about the GBO being shut down...

We need to rethink all this. Especially for Ostap (and future users) who are not supervisors. One solution is to provide independent power. We could tap the continuous Observing Pad circuit and put one outlet in the Warm Room (for the switch) and one in the Observing Room (for the dehumidifier and WAP)... Would solve the problem. In the meantime, more extension cords will do that trick. Another idea is we just leave the GBO powered.

Dietmar said Ostap won't need it for a while...