How very odd.
It used to work. Now it doesn’t.
Trying to log onto https://www.schwab.com from my Raspberry Pi or Odroid XU4 fails to even open a window. Just hangs with a blank screen.
I have no idea why.
It does work with FireFox on the Mac, but it is running an older FireFox as it can’t take updates anymore. I don’t have other browsers on the Linux boxes, well, at least none that I’d say works… so can’t really debug it more than that. I’ve also logged on via the tablet in the past, so it does work with Android browsers and on an ARM based CPU.
Why would it suddenly stop working? (Well, really, stopped about 2 months ago, but I figured the next update would “fix” it. It didn’t.)
The Mac is running through the same routers and all, so not a network thing. It’s got to be something specific to Firefox on Linux (or on these particular flavors of Linux. One Devuan the other Armbian).
I suppose sometime “later” I’ll try booting up an old Ubuntu or Red Hat Pi image and see if they work. I don’t really want to swap platforms for that one use, but I can easily have a “dedicated Schwab trading chip”.
I do wonder about the timing as it is right in sync with the FireFox “upgrades” to deal with Meltdown and Specter bugs. I did notice some familiar pages loading and running a bit slower on the XU4 than prior to the “fix”.
FWIW, the Birthday Bash brought to me 2 more SBCs. My cluster is now made up of 2 x Pi M2, 2 x Pi M3, 2 x Orange Pi One, and an Odroid C1. (I’ve moved the XU4 over to desktop useage, but it can join the cluster if desired). I’ve spent the last couple of days rearranging the desk and getting the cluster bits all together. I still need to clean up the network a bit. Found my old Netgear 8 port switch, but not the power brick. It’s in a box… somewhere… so eventually I’ll have that 100 Mb switch in place. For now, I stuck an old hub on the Netgear Router as a port addition.
So that’s basically 16 cores of Pi, 8 of Orange, and 4 of Odroid (with an additional 8 if desired). Total of 28 in the basic config and 36 “all up”.
I presently have most of it warming up on Boinc, but the two newest boards are not playing well with it yet. I’ve consulted my old Boinc posting, but there’s still something not quite right. They all communicate, but the Boinc Manager isn’t talking to them. Most likely one of those config / permissions settings things. I also need to finish the install / configure of “distcc” on the new members, then try some big build on it.
At this point, I have more computes than I’m presently able to effectively use. Guess it’s time to get back at that parallel FORTRAN and Models project ;-) and that learning Julia… I’d use them to mine crypto-coins but their small core size makes it kind of silly and I’m just not really “into” the whole digital currency thing anyway. Besides, the whole field has moved off to dedicated ASIC / hardware land. I’d be mining with a teaspoon while everyone else was using power shovels and blasters.
Well, that’s the status update on the cluster stuff for now. Back to debugging, configuring, and testing things.