Ac5000

joined 1 year ago
[–] [email protected] 1 points 1 year ago

I don't think the bifurcation is causing me issues. Before I enabled it, I wasn't able to see the drives from iDRAC/Bios. From what I've been able to research, this is expected and Dell sells the "solution" to booting directly from them. (Add-in card that's pretty pricey...)

I do have an old SATA SSD that I'm considering slotting into one of the bays and using to boot. But I see that as a "last resort" option. I was hoping to have a bit of redundancy with the Proxmox install/configuration itself.

I feel that there's a solution to the current setup and I just lack the knowledge to fix it. Everything I've been able to find points to my current setup being able to work. I'm just being hindered by not understanding partition/device/boot structure.

From what I understand, and what I saw during the Proxmox installation, if I can get past whatever part of the POST/boot process is preventing seeing the drives directly, I can use Clover to bootload from there. I've been able to boot into Clover just fine, and it was able to "see" the drives and partitions. I just don't know which one should hold the Proxmox boot and if I've configured the Clover config correct.

[–] [email protected] 1 points 1 year ago

From what I've read online, Dell does something similar. There's some sort of card/add-on that can enable directly seeing and booting from PCIe but they are costly.

This server has the internal USB and a build in SD slot accessible from the rear. (There's also a dual card option like you mention for redundancy.)

My plan was to get Clover working with USB, then use the vFlash SD slot to hold the Clover bootloader in read-only mode. This would hopefully prevent the SD card from dying quickly.

[–] [email protected] 3 points 1 year ago* (last edited 1 year ago)

Thank you for posting this with the explanations and great visuals! I am wanting to upgrade to a setup almost identical to this and you've basically given me the bill of materials and task list.

Anything you wish you had done differently or suggest changing/upgrading before I think about putting something similar together?

[–] [email protected] 1 points 1 year ago

Any way you could update/create your own drawing with what you mean? (Bad paint drawings are acceptable!)

I ask because I am curious if I am subject to the same problem. I'm not the most networking savvy so I need the extra help/explanation and maybe the drawing will help others.

[–] [email protected] 5 points 1 year ago

I'll throw my code into the ring as well. I posted it over in the Python community and have been using it myself.

It's not the most user friendly yet though. Still working on improving it as I get time though and open to suggestions/requests.

https://github.com/Ac5000/lemmy_account_sync

https://lemm.ee/post/608605

[–] [email protected] 3 points 1 year ago (1 children)

Sorry for the delay in getting back.

Currently it will not work with 2FA enabled. However, looking at the login post requirements I just need to add that as an option to put in the config.

I'll reply to this comment again when I get something put together. I'll add it to the GitHub issues list for tracking as well.

However, could you recommend an instance that uses 2FA for login so I can make an account to test it? I see the field in my current instances but would like something fresh to try it on.

[–] [email protected] 2 points 1 year ago (1 children)

Someone already posted the wescode migrate, but here's my Python code that does a bit more. It's not the most user friendly yet, but I'm working on it.

https://github.com/Ac5000/lemmy_account_sync

[–] [email protected] 2 points 1 year ago

Correct. lemmy.world was my original account. But with the server strain going on, I've hopped over to lemm.ee and also have a couple other accounts. Just run this script every so often and all your accounts will more or less feel like the same account.

[–] [email protected] 1 points 1 year ago* (last edited 1 year ago)

Yeah, for anyone that gets the 502 gateway error. That means the instance was down when it tried to login/didn't respond. I'm going to revisit this part of the code later and see if I can fix/handle that happening so it at least goes through with the rest of the accounts. If it happens to you, just run it again and hopefully you'll get through at least once.

Subsequent runs actually hit the servers with less requests since you can pull most of the info you need with the initial site response and I check that before making any needed requests.

Also, someone else mentioned they had a problem with the none type thing. I'm using Python 3.11 and forgot to specify that. I'll add it to the readme when I get a chance.

view more: ‹ prev next ›