Conversation

Notices

1.  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 04:22:40 JST 
   >booted arch iso on a spare pc to bootstrap debian from it
   >kernel panic when copying files over
   >after hard reboot 4 gigs of memory are once again gone
   :niggainsane:
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 06:00:06 JST 

     in reply to
     Wonder if anything bad would happen if I just yeet the APU's vBIOS from the BIOS update image.
     Screenshot_20230731_235851.png
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 06:09:09 JST 

     in reply to
     It won't let me flash it, probably some signature issues or something.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 06:37:19 JST 

     in reply to
     Clearing CMOS didn't help this time, either.
   • Vivi Nella Verita :Debian_logo: (verita84@poster.place)'s status on Tuesday, 01-Aug-2023 06:39:09 JST Vivi Nella Verita :Debian_logo:

     in reply to
     @mint
     
     Hey that's how I install Debian lol
      likes this.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 08:23:24 JST 

     in reply to
     Might be bad or just incompatible RAM. I swapped sticks around and the system stopped booting entirely. BIOS and memtest still detected all four sticks though.
   • Ethical Pedophile (blinkrape@posting.lolicon.rocks)'s status on Tuesday, 01-Aug-2023 08:28:08 JST Ethical Pedophile

     in reply to
     @mint blow the RAM sockets out with compressed air. Could be a flaky port or flaky stick tho. only way to know is test them one at a time.
      likes this.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 08:30:14 JST 

     in reply to

     • Ethical Pedophile
     @BlinkRape Will do. Worth noting there's no speaker beeps or anything, only spinning fans and some power going to PCI/SATA.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 08:44:15 JST 

     in reply to

     • Ethical Pedophile
     @BlinkRape Tested two random sticks in a first slot, the system didn't boot with either of them. I've ran memtest on remaining 12 gigs before and it passed, so it probably isn't an issue with the sticks themselves.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 08:57:11 JST 

     in reply to

     • Ethical Pedophile
     @BlinkRape Got the system running again by plugging back all sticks and reset the CMOS a thousandth time. Now it's booting but sees only 8 gigs. :heyface:
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 09:04:01 JST 

     in reply to

     • PC-9801 Enjoyer
     @Pawlicker Maybe a clocking issue? That's the same RAM which was running flawlessly in my desktop before upgrade; It's running at 1866 MHz now with no way to change it in BIOS, but I'm pretty sure it was at 1600 MHz before.
   • PC-9801 Enjoyer (pawlicker@bae.st)'s status on Tuesday, 01-Aug-2023 09:04:02 JST PC-9801 Enjoyer

     in reply to
     @mint bad ram classic sign imo
   • Ethical Pedophile (blinkrape@posting.lolicon.rocks)'s status on Tuesday, 01-Aug-2023 18:17:37 JST Ethical Pedophile

     in reply to
     @mint remove one stick at a time and see if you find the problem socket. then swap other RAM into that socket and see if the problem persists. If not you've found your faulty module.
     
     You could also run an overnight memtest86 scan on all of them while its on I guess.
      likes this.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 19:08:44 JST 

     in reply to
     Looks like one stick is bad, after all. Tried all four sticks in a single slot, with one of them system started beeping (three short, one long) the same way as if there's no RAM installed at all.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 19:10:06 JST 

     in reply to
     Sometimes there's beeps, sometimes it just freezes on boot.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 19:17:12 JST 

     in reply to
     And when I put it *together* with an older stick of 1600 memory, it suddenly boots and detects again. Doing the memtest now.
   • gray (gray@ryona.agency)'s status on Tuesday, 01-Aug-2023 19:28:56 JST gray

     in reply to
     @mint I got a stick of Kingston RAM for my Chinese mini PC and had a similar problem with it. Sometimes it’s POST and load the OS, sometimes it’s POST then crash, and sometimes would error beep. I never got around to running memtest but it was so weirdly inconsistent with the faults it showed.
      likes this.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 19:41:58 JST 

     in reply to

     • gray
     @gray Memtest on 1600 MHz passed. I've put back the other 1866 stick, and it's back at only detecting 4 gigs. At this point I might as well roll with 14 total gigs until I get a matching 1600 MHz 4 gig stick.
   • gray (gray@ryona.agency)'s status on Tuesday, 01-Aug-2023 20:06:14 JST gray

     in reply to
     @mint I’d have to look at the specs of that Kingston RAM but I think it’s DDR4 1866 and the system can only do 1600. I thought it’d just run at the lower speed and not just shit the bed.
      likes this.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 20:07:19 JST 

     in reply to

     • gray
     @gray No settings in BIOS to change the speed? Same case with my chinkcentre, so I had to downgrade it a bit now.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 20:08:59 JST 

     in reply to

     • gray
     @gray But the stick might be bad, after all. All three other sticks have no problem running at 1866 (or underclocking them to 1600, idk). All sticks came from the same bundle but the problematic one has the earliest year and week of production, so maybe it's running with older and buggier microcode.
   • gray (gray@ryona.agency)'s status on Tuesday, 01-Aug-2023 21:06:52 JST gray

     in reply to
     @mint those chink mini pc don’t have shit in their BIOS
      likes this.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 21:07:40 JST 

     in reply to

     • gray
     @gray Both of mine have quite a few options, though I don't remember if there were a nothing about clock speeds.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 21:10:06 JST 

     in reply to

     • gray
     @gray Just took a look at RAM prices, and damn it's getting pretty cheap. 8GB DDR3 DIMM is less than a thousand rubles. I'll try getting two of them and seeing if I can get it running at the whole 24 gigs.
   • gray (gray@ryona.agency)'s status on Tuesday, 01-Aug-2023 21:12:50 JST gray

     in reply to
     @mint last I looked in the settings on mine it pretty much just had UEFI settings and secure boot settings and that’s it.
      likes this.
   •  (mint@ryona.agency)'s status on Tuesday, 01-Aug-2023 23:30:51 JST 

     in reply to

     • gray
     @gray Sticks obtained, the system detected 24 total gigs just fine, taking memtests now. Took a detoure to get some import sodas and was immensely disappointed as there were no vanilla Dr Pepper or lime Schweppes.
   •  (mint@ryona.agency)'s status on Wednesday, 02-Aug-2023 00:35:15 JST 

     in reply to
     Memtest passed.
   •  (mint@ryona.agency)'s status on Wednesday, 02-Aug-2023 04:16:51 JST 

     in reply to
     When setting up the server, noticed it gets a proper IPv6 with a delay. Started debugging this issue with other device I have on hand and noticed Android doesn't get it at all. Apparently, when you're using radvd, putting a subnet larget than /64 makes the clients spaz out.
   •  (mint@ryona.agency)'s status on Wednesday, 02-Aug-2023 04:19:58 JST 

     in reply to

     • gray
     @gray Yeah, that's also why I was confused as I was getting an address just fine via DHCPv6.
   • gray (gray@ryona.agency)'s status on Wednesday, 02-Aug-2023 04:19:59 JST gray

     in reply to
     @mint Doesn't Android only use SLAAC too?
   • gray (gray@ryona.agency)'s status on Wednesday, 02-Aug-2023 04:22:57 JST gray

     in reply to
     @mint It would be nice if ipv6 would just work
      likes this.
   •  (mint@ryona.agency)'s status on Wednesday, 02-Aug-2023 05:31:11 JST 

     in reply to
     Alright, basic system is set up the way I wanted, and there's plenty of room for activities. In following days I'll probably set up Wireguard, LXC and try moving Agency there.
     Screenshot_20230801_232437.png
   •  (mint@ryona.agency)'s status on Wednesday, 02-Aug-2023 05:34:15 JST 

     in reply to

     • gray
     @gray Right now, yes. The rest of the services would be in Alpine containers as I like.
   • gray (gray@ryona.agency)'s status on Wednesday, 02-Aug-2023 05:34:16 JST gray

     in reply to
     @mint Are you just using a barebones Debian install?
   •  (mint@ryona.agency)'s status on Wednesday, 02-Aug-2023 05:43:11 JST 

     in reply to

     • gray
     @gray Haven't bothered checking it out, and since the amount of container would be in single digits, it's probably not worth the effort when I can just ssh into the machine and manually set everything up as needed.
   • gray (gray@ryona.agency)'s status on Wednesday, 02-Aug-2023 05:43:12 JST gray

     in reply to
     @mint is there a reason you're not using something like Proxmox for containers? Just curious.
   •  (mint@ryona.agency)'s status on Wednesday, 02-Aug-2023 06:05:03 JST 

     in reply to
     Ok, maybe not pure Wireguard. Seeing reports of it getting blocked, and another service I host (with Wireguard tunnel to OVH VPS) also stopped working a few hours ago.
     https://ntc.party/t/wireguard/4968
   • laurel (laurel@freespeechextremist.com)'s status on Wednesday, 02-Aug-2023 07:06:55 JST laurel

     in reply to
     @mint
     
     >blocking VPNs
     How does this work for business in Russia that use VPNs as part of their operation?
     Is the двач /s/ thread a good place to be informed about developments on centralized blocking in Russia?
     Speed considerations aside, can't people use an ssh tunnel? Surely they can't start blocking those without their own infrastructure falling apart.
      likes this.
   •  (mint@ryona.agency)'s status on Wednesday, 02-Aug-2023 07:10:29 JST 

     in reply to

     • laurel
     @laurel SSH is as easy to block as Wireguard, and you generally don't transfer huge amount of data at random intervals through it. Right now it all seems selective, but your best bet would be looking what tunnels Chinese use. I have two Shadowsocks servers running, both wrapped in v2ray, and a backup dnstt node, but haven't used any of them for a while since my current VPN provider isn't blocked yet. As for Wireguard on OVH, wrapping it in udp2raw in faketcp mode helped.
   •  (mint@ryona.agency)'s status on Wednesday, 02-Aug-2023 16:39:26 JST 

     in reply to

     • laurel
     @laurel No keys in FSB's hands, no VPN. Simple as.
   • laurel (laurel@freespeechextremist.com)'s status on Wednesday, 02-Aug-2023 16:39:27 JST laurel

     in reply to
     I can understand blocking popular VPN providers. But blocking VPN software, either openvpn or wireguard, will cause problems with firms using this kind of software for legitimate uses. Even more so if someone tries to block ssh.
     
     I've looked a bit into shadowsocks and dnstt before and I'm sure they work much better than ssh tunnels.
     But it's really about obfuscating them, like the http obfuscation of v2ray, or GoQuiet shadowsocks addon, or routing everything through obfsproxy.
     Machismo repeated this.
   • laurel (laurel@freespeechextremist.com)'s status on Friday, 11-Aug-2023 02:10:46 JST laurel

     in reply to
     @mint
     
     >but your best bet would be looking what tunnels Chinese use.
     I looked a bit into this and found something that you may find interesting.
     It's a pretty good write up on how the Chinese firewall works, pretty recent too (April 2023).
     I found it very interesting that they ignore UDP traffic completely.
     https://gfw.report/publications/usenixsecurity23/en/
      likes this.
   •  (mint@ryona.agency)'s status on Friday, 11-Aug-2023 02:16:16 JST 

     in reply to

     • laurel
     @laurel There's been a few articles on Habr about block circumvention.
     https://habr.com/ru/articles/727868/
     https://habr.com/ru/articles/731608/
     Apparently, the current meta is XTLS-Vision masquerading as a legitimate site, and maybe VLESS-over-gRPC (which can be wrapped into some CDN like :cloudflare:) as well.