|
Jonny 290 posted:Let's do a detective thing. not sure on the cause of initial incident, but as you were plugging things back together is it possible the Debian box with all of its VMs had a lease that your Xfinity modem handed out? Not sure if the scopes of the consumer stuff were the same but that would be the first thing I'd look at
|
#
?
Jan 12, 2022 21:11
|
|
|
|
| # ? May 19, 2024 09:57 |
|
|
Right, but even if the lease scopes were different (yes), why would, from my desktop that wasn't being touched, i suddenly stopped being able to ping 8.8.8.8 as soon as i plugged the debian box in? The xfinity router, as lovely as it is, would never hand out the gateway IP as a lease. which would be the only thing that would break that.
|
|
#
?
Jan 12, 2022 21:28
|
|
|
the scenario i'm thinking of is your machine which you plugged into the router got handed a lease that conflicted with something on that other box. Wouldn't be an issue until both are plugged in simultaneously. if you had an IP address conflict between your pinging device and something on the other box, you'd experience intermittent connectivity.
|
|
#
?
Jan 12, 2022 21:38
|
|
|
Mmm, yeah. Everything's internally natted on the debian box, it got 10.0.0.12 when i turned its monitor on (again.  ) and my desktop was .6.
|
|
#
?
Jan 12, 2022 21:40
|
|
|
If there isn't anything in the logs on the Linux box, my best guess is a low level bitflip on the NIC caused it to send just the right nonsense (like in handshaking/QoS stuff) to cause the switch to lock up.
|
|
#
?
Jan 12, 2022 22:14
|
|
|
Literally could have been the result of a solar flare OP.
|
|
#
?
Jan 12, 2022 22:14
|
|
|
Yeah it's a 1 in a billion edge case and i'm mad at myself for power cycling the server before looking at its status. It's just the weirdest loving networking thing i've ever seen in my life, and i had a ccna cert. i cannot explain it beyond arp table horseshit, but the switch is very dumb and very reliable, and it's all just so....what
|
|
#
?
Jan 12, 2022 22:16
|
|
|
yeah that's weird. so are you keeping the xfinity router in there for now?
|
|
#
?
Jan 12, 2022 22:27
|
|
|
If nothing is in the logs it's probable the OS never got a notification that anything was wrong. Presumably you had software on the box that would have been trying to use the NIC, and if the driver received notice of a fault it should have been logged. If it happens again, check if the firmware package for the NIC had a recent update.
|
|
#
?
Jan 12, 2022 22:31
|
|
|
Hed posted:yeah that's weird. till the weekend, i don't wanna gently caress up my connection during work hours, and i want my plex for after work SYSV Fanfic posted:If nothing is in the logs it's probable the OS never got a notification that anything was wrong. Presumably you had software on the box that would have been trying to use the NIC, and if the driver received notice of a fault it should have been logged. Yeah, good idea, will do.
|
|
#
?
Jan 13, 2022 18:40
|
|
|
my second monitor - an asus vg27aq - goes into soft power off every time i stand up or sometimes just of its own accord and sometimes it just decides to blank itself for a moment, and the very top flickers red before going completely dark other than that it's acting like the monitor i paid for, four stars  updated card drivers, switched e: hdmi is not dp flakeloaf fucked around with this message at 20:51 on Jan 19, 2022 |
|
#
?
Jan 19, 2022 18:43
|
|
|
Do you sit in an office chair with a piston on it? Some DisplayPort monitors shut off when an office chair is operated nearby. No, I am not pulling your leg. https://www.youtube.com/watch?v=voW5kEI7JKE It's an EMI shielding problem and the solution is to try running your cables differently, or get a different chair. https://support.displaylink.com/knowledgebase/articles/738618-display-intermittently-blanking-flickering-or-los Sagebrush fucked around with this message at 18:50 on Jan 19, 2022 |
|
#
?
Jan 19, 2022 18:45
|
|
|
Sagebrush posted:Do you sit in an office chair with a piston on it? Some DisplayPort monitors shut off when an office chair is operated nearby. well innat somethin efb i do, in fact, and the cloth surface generates a hell of a zap when i get up or move... and the intermittent flickers tend to happen when i look over at the other screen, which i can't do without shifting in my chair, and i am definitely not replacing the chair so it's shieldin' time thanks!
|
|
#
?
Jan 19, 2022 18:54
|
|
|
some ferrite cores on the cable ends are often effective
|
|
#
?
Jan 19, 2022 19:20
|
|
|
in a similar vein, I had a fun problem it took me a while to identify; static build-up and a metal desk frame meant any, even mild, zap was enough to trip my mouse into waking my desktop. it's conditional on humidity, so it would come and go seasonally, sometimes daily in fall and spring I'm fairly certain it's the cable in conjunction with the mouse, as it stops if I unplug the mouse-end of the cable (it's a trackball with a usb a to usb micro cable) so I need to do what jonny suggests it doesn't help that I routed my cables out of view by running them parallel in a bundle to the desk frame at the back, so any zaps on the desk or through a different cable is enough Agile Vector fucked around with this message at 19:45 on Jan 19, 2022 |
|
#
?
Jan 19, 2022 19:43
|
|
|
I once upgraded a 3D printer with a new faster control board, and from then on it would shut down about 30 seconds after starting a print. The point at which it would shut down was different from print to print, but almost identical every time I would retry any given file. I spent literally weeks digging into every possible combination of hardware and software, and eventually traced it down to the stepper motor driver boards getting reset for some reason. After over a month I had the whole thing pulled apart with multiple data lines connected to an oscilloscope, decoding the driver commands and trying to figure out what the gently caress was going on and still seeing no reason for it to fail. Frustrated, I was idly pushing the extruder carriage back and forth, and suddenly the oscilloscope triggered on an 80 volt spike. In the end it turned out something about the physical configuration of the extruder carriage -- I decided it was either the plastic rollers rubbing on the frame, or the drive belt acting like a Van De Graaff generator -- was building up static electricity from motion, which would reach about 80 volts and then discharge through some route I never found into the control circuitry, resetting the drivers. Presumably it had been doing this all along, and the older control board just wasn't as sensitive. I ran a ground wire from the carriage frame to the power supply and instantly everything worked perfectly. 
|
|
#
?
Jan 19, 2022 20:07
|
|
|
^that is completely bonkers, and the way you discovered it is peak scienceAgile Vector posted:in a similar vein, I had a fun problem it took me a while to identify; static build-up and a metal desk frame meant any, even mild, zap was enough to trip my mouse into waking my desktop. it's conditional on humidity, so it would come and go seasonally, sometimes daily in fall and spring powercfg /lastwake blames my case's power button for half of my phantom wakeups which, hey hey, happen right after i get up that must be some emp flakeloaf fucked around with this message at 20:12 on Jan 19, 2022 |
|
#
?
Jan 19, 2022 20:10
|
|
|
Make sure you are using a cable that came in a box with the Display port or HDMI logo on it from a reputable retailer (shipped/sold by amazon cables usually pass verification). Until you can get a new cable that's guaranteed to meet spec , try disabling any adaptive sync if it's enabled. Try a different surge suppressor, and if your surge suppressor has relays in it, make sure the monitor is not powered through a port with a relay. Double check that updates haven't reverted your driver version to something ancient. A little bit of strain over a long enough time will damage the cable, some monitors do not handle adaptive sync well, and HDMI/Displayport can be very sensitive to electical jitter. Edit: if you've got one of those sensing power strips, you can get corrosion on the relay contacts. Taking it apart and using contact cleaner on it will usually fix it. SYSV Fanfic fucked around with this message at 20:20 on Jan 19, 2022 |
|
#
?
Jan 19, 2022 20:13
|
|
|
heh that's funny, cause i changed the cable out first for one of these during troubleshooting, and that changed the complaint from "it often flickers" into "sometimes it just shuts off". That made me suspicious that it was a static something, because a cable's just an antenna that two people think they aren't sharing, but i didn't get the connection between the outages and my rear end until i shut the light off, THEN got up, and noticed everything behind me got dark for no reason it's plugged into the same UPS the computer is (an APC back-ups 1500), but I see it's in the "MASTER" slot, next to one labelled "CONTROLLED BY MASTER"... so yeah, there's probably a relay in the vicinity. I've moved it to one of the battery ones instead; good thought 
|
|
#
?
Jan 19, 2022 20:27
|
|
|
My laser printer is on the same circuit as my 13.8v 50 amp linear power supply for my ham gear. If the power supply switch is on, the printer goes full pudding brain and prints about 3 lines of the first page of anything I send, and then just cycles the whole paper tray contents through it. Power supply off, prints perfectly. My suspicion is that on startup, the laser printer and extremely large transformer in the ham power supply interact in a deleterious fashion, and it fritzes out its internal cpu or whatever.
|
|
#
?
Jan 19, 2022 20:41
|
|
|
HDMI is a really high throughput link - the latest version goes up to 48gbps. Even 1080p is like 5gbps. No room for jitter at those speeds.
|
|
#
?
Jan 19, 2022 20:44
|
|
|
whoops yeah i meant dp the whole time; i tried hdmi on a pc monitor and it was miserableJonny 290 posted:My laser printer is on the same circuit as my 13.8v 50 amp linear power supply for my ham gear. If the power supply switch is on, the printer goes full pudding brain and prints about 3 lines of the first page of anything I send, and then just cycles the whole paper tray contents through it. Power supply off, prints perfectly. tempest fujitsu
|
|
#
?
Jan 19, 2022 20:51
|
|
|
Jonny 290 posted:My laser printer is on the same circuit as my 13.8v 50 amp linear power supply for my ham gear. If the power supply switch is on, the printer goes full pudding brain and prints about 3 lines of the first page of anything I send, and then just cycles the whole paper tray contents through it. Power supply off, prints perfectly. That would greatly concern me if I cared about the printer. Most likely the radio is f'n up the 60hz (50hz in pal territory) waveform from the power company,, and it's confusing a switching power supply in the printer. If the voltage gets low/high enough it's going to flip values in ram. .
|
|
#
?
Jan 19, 2022 20:57
|
|
|
You can test this with a cheap plug in digital clock if you don't have a meter than can tell you. Most of them don't have a crystal and keep time off the oscillation of the power grid. The time will drift over the course of a day or two.
|
|
#
?
Jan 19, 2022 21:00
|
|
|
SYSV Fanfic posted:That would greatly concern me if I cared about the printer. Most likely the radio is f'n up the 60hz (50hz in pal territory) waveform from the power company,, and it's confusing a switching power supply in the printer. If the voltage gets low/high enough it's going to flip values in ram. . Yeah i am thinking about moving it to another circuit on the other branch. lasers have that big current spike on startup, and i think im getting a little bit of inductive kickback somehow
|
|
#
?
Jan 19, 2022 22:09
|
|
|
are you sure it's actually the laser that's drawing all the current and not, like, the fuser
|
|
#
?
Jan 19, 2022 22:16
|
|
|
jonny's probably saying "lasers" as a shorthand for "laser printers, which include the fuser component" and not literally "lasers, like as found in 'laser printers', are high-current devices"
|
|
#
?
Jan 19, 2022 22:28
|
|
|
correct
|
|
#
?
Jan 19, 2022 22:58
|
|
|
flakeloaf posted:powercfg /lastwake blames my case's power button for half of my phantom wakeups which, hey hey, happen right after i get up yeah, I think that's what lead me to single out the mouse. for a bit I thought it was just that sensitive to me bumping the desk until I got a strong static shock when I brushed the edge kicking it off by standing is next level though! flakeloaf posted:tempest fujitsu  Farmer Crack-rear end posted:jonny's probably saying "lasers" as a shorthand for "laser printers, which include the fuser component" and not literally "lasers, like as found in 'laser printers', are high-current devices" well that's not as cool
|
|
#
?
Jan 19, 2022 23:44
|
|
|
crappy update: moving the cable and switching it to an outlet without a relay accomplished nothing, so it's time for this thing:Jonny 290 posted:some ferrite cores on the cable ends are often effective apparently fit matters, so i'll stick a vernier on this cable and try to find something with a roughly-equivalent ID
|
|
#
?
Jan 20, 2022 17:20
|
|
|
not really sure if theres a better place to post this, but does anyone know if its possible to display a 'RECAP' for an ad-hoc ansible command? eg: ____________ < PLAY RECAP > ------------ localhost : ok=18 changed=8 unreachable=0 failed=0 skipped=4 rescued=0 ignored=2
|
|
#
?
Jan 21, 2022 00:17
|
|
|
epilogue: 7mm ID ferrite rings are too small for a 7mm OD cable, and if you persist, the rings will break first secretlabs responded to my email with a recommendation for humidfiers; i think we can assume they did not read it e: they also recommended i put the chair in another room. this wouldn't make it less staticky, but it would prevent me from touching computer 
flakeloaf fucked around with this message at 03:16 on Jan 22, 2022 |
|
#
?
Jan 22, 2022 03:08
|
|
|
while i can vouch for, uh, humidity, for my problems, i think 'sit away from your computer' is the best customer service reply i've seen in a while
|
|
#
?
Jan 22, 2022 04:52
|
|
|
flakeloaf posted:epilogue: 7mm ID ferrite rings are too small for a 7mm OD cable, and if you persist, the rings will break first humidity is an important part of electrostatic mitigation. esd sensitive work areas generally have minimum humidity requirements
|
|
#
?
Jan 22, 2022 08:01
|
|
|
oh absolutely, and the problem's definitely worse at 35% than it was at 70% (RH at 17-22C), but it was still there at 70 and, past that point, my area is less of a toybox and more of a terrarium
|
|
#
?
Jan 22, 2022 16:30
|
|
|
flakeloaf posted:oh absolutely, and the problem's definitely worse at 35% than it was at 70% (RH at 17-22C), but it was still there at 70 and, past that point, my area is less of a toybox and more of a terrarium make the best of it and put some tropical geckos in there
|
|
#
?
Jan 22, 2022 16:33
|
|
|
flakeloaf posted:oh absolutely, and the problem's definitely worse at 35% than it was at 70% (RH at 17-22C), but it was still there at 70 and, past that point, my area is less of a toybox and more of a terrarium wow, yeah, above 40 is the cut-off for my static issues. still happening at 70 is a lot Jenny Agutter posted:make the best of it and put some tropical geckos in there 
|
|
#
?
Jan 22, 2022 17:29
|
|
|
my landlady has hosed up our fiber internet and the best we can now get is basic adsl. she is suggesting getting a 4g/5g enabled router with unlimited data instead. how bad an idea is this?
|
|
#
?
Jan 27, 2022 20:49
|
|
|
vodkat posted:my landlady has hosed up our fiber internet and the best we can now get is basic adsl. she is suggesting getting a 4g/5g enabled router with unlimited data instead. how bad an idea is this? Definitely not fiber speeds but a buddy of mine has t-mobile's Nokia made 5G wastebin looking thing and he consistently gets 20-50mbps symmetrical out in the middle of nowhere.
|
|
#
?
Jan 27, 2022 20:58
|
|
|
|
| # ? May 19, 2024 09:57 |
|
|
how pray tell does a landlord "gently caress up" fiber internet like it worked before and its broken now? and cannot be repaired?
|
|
#
?
Jan 27, 2022 21:09
|
|
