• 4 Posts
  • 66 Comments
Joined 3 years ago
cake
Cake day: July 2nd, 2023

help-circle

  • brewery@lemmy.worldtoSelfhosted@lemmy.worldWhy homelab?
    link
    fedilink
    English
    arrow-up
    8
    ·
    11 days ago

    When you have nextcloud, immich, vaultwarden and navidrome, this is real data that you’re talking about on services that I use several times a day. I had an overnight blackout that took me a few hours to recover from. No data lost and have a 3-2-1 backup system but that takes time and energy to fully recover from.

    After that, bought the cheapest UPS I could find, installed NUTS and not had issues since. Most blackouts seem to happen overnight and all have been less than 5 minutes.

    Buying a UPS that would last for 12 hours minimum (happens just after I go to sleep and I realise in the morning or happens at work and not back home until then) is much more expensive and difficult than installing CUPS to have it switch off. Then I can wake up, turn the machine back on, wait a few minutes and everything will be back up. I can run tests but if I need to go to work I can leave that to later and what works I can use.


  • Worked in the UK for one of the largest Japanese multinationals. Our team required several government accounts with different passwords. I asked IT what are my options and they said we have none, and that their team used a text file on a shared drive so I did it on Excel instead to make to make it more searchable. If they’re not willing to let you install keepass or buy any software, what can you do!


  • I found the Cisco free networking courses really helpful when trying to figure out my network stack. Some of the earlier stuff you probably know but you can skip through quickly, especially if you don’t care about getting the “qualification”. https://www.cisco.com/site/us/en/learn/training-certifications/training/netacad/index.html

    I would personally recommend keeping the router and adguard/pihole server separate. If you mess something up, it’s easier to quickly switch the router DNS to a public one whilst you try to fix it, especially if you have anybody else at home who won’t appreciate losing their internet because you messed up something (which will happen but let’s you learn)! It’s also a good idea to keep this separate from your main server for this very same point. You might accidentally use up the entire hard drive or have some messed up update on one software take all the RAM, making everything else unusable (both happened to me at different times, and one was overnight before I was working from home which was not fun to try and fix first thing in the morning!). It’s not always clear which device is causing and what went wrong so can take some time to fix but gives you flexibility to fix in your own time.

    That’s why I would recommend getting a used thin client for £30ish (GBP) from eBay or somewhere like that. I have my pihole and home assistant on there to be clean and separate in case my server goes down. Then the internet still works, the lights still work, and things like nextcloud and jellyfin just have to wait a bit.

    It also allows you to upgrade devices separately (e.g. when my internet provide unexpectedly upgraded to 2.5G I chose to buy a new router and one of the switches, and upgrade the Ethernet socket on my home server) but kept my existing WiFi access point and pihole/HASS server which still work fine after years of use. Most of my clients at home can’t handle higher speeds so no point upgrading the rest of the network until they can (if ever). I have separated my WiFi access point and router because of the location of my incoming wires and home layout, which you could also think about doing. I’ve always thought getting seperate devices focusing on one thing they do best makes sense, and can actually be cheaper to buy.

    Not sure about routers themselves but have heard good things about openwrt. I do NOT recommend the TP Link Omada series!







  • Given that you can also export the vault in json, CSV and json (encrypted), it should not be too difficult to transfer to the new one. I have migrated from keepass to vaultwarden, then moved the vaultwarden between home servers without difficulty.

    I backup vaultwarden locally and to borgbase, and managed to get my vault backup after simulating a crash. However in reality, the few times I have had actual problems with a connection while away from home, I made do with the local copies available on my phone, tablet and/or laptop. You can’t create new passwords if your Server is down but you have access to the client copy of existing passwords since it last synced. I don’t keep my email password on there so in one rare case I just created a new password to logon to something urgently.


  • My manager and a team mate had to go to the US for work earlier this week. They’re both British citizens but Indian heritage so was genuinely worried for them. Work made sure their papers were in order and they had local contacts just in case. It was “no more difficult than usual” (it’s always been hard for brown people coming in with “random” security checks). There is no way I would be going though, would flat out refuse.

    One made it home but funnily enough the other is stuck in Vancouver on the connection as theres some fire in Heathrow. At least he’s out of the USA!




  • Afraid not, was just flicking through and saw Raleigh in the UK section, which I know a lot about as lived in the area of Nottingham it was once in. They offshored all their factories to cheaper Asian countries many years ago leaving just a small presence in the UK to sell and unpack the bikes. I can see only 130 UK staff (68 sales, 62 admin) so can’t imagine they build anything here.

    When I get some time I’ll go through the list, at least the UK ones which I can find information about much more easily (our company data is open to all - not many countries have this)




  • Cant really talk about product quality but from a social perspective, which I can see pretty clearly given I have worked for a major Japanese conglomerate and spent a lot of time with Japanese “expats” across Europe, it’s very far behind Europe on social aspects.

    There is a lot of misogyny, racism, hierarchy, classism and conservatism. Theres no recognition, acceptance or real change done in relation to the horrors carried out in WW2.

    Part of the reason for their very low birth rate is their poor treatment of working women and extreme tough stance on immigration or refugees. Their working culture is horrendous generally.

    Young people are pushing back and trying to change things so if it’s a small, modern company they may be OK. However, if you care for social issues then avoid the major conglomerates that run most of the economy.

    On the plus side, they are very good at technical engineering and really care about their (Japanese) workers, which still have jobs for life.




  • I was thinking about this last night. In wars we often issue war bonds and people go around asking everyone to contribute what they can. Why not reinvent this but concepts including co-operatives, employee councils, kickstarter, social awareness, community awareness, inclusion of minorities and socially disadvantaged people, etc.

    How about a technology investment bond, or a direct investment fund into new enterprises to build these up.

    Shame the rich to invest into them instead but we know that won’t work so let’s all invest so we don’t have to rely on them.

    I know it’s politically challenging but I would happily invest my pension pot into companies to improve our European tech in a socially aware and fair system, than the current US stock market it’s half made up of (I will change this!). Hell, just give grants to the open source community to develop things and let everyone in the world benefit