Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


What do people do with the cheap Kimsufi servers? KS1 & KS-LE-1?
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

What do people do with the cheap Kimsufi servers? KS1 & KS-LE-1?

stonedstoned Member
edited November 2022 in General

Hi guys, happy thanksgiving and holiday season to everyone! Thanks for a great forum and deals!

I purchased this BF2022:

  • 2x KS1
  • 1x KS-LE-1
  • 2TB storage VPS for $3/mo with Hosthatch - $110 for 3 years

I bought 2x KS1 for the 2x 2TB for total of 4TB storage, but these are single disk systems, with no redundancy.

The KS-LE-1 is nice. Older CPU, 32GB ECC RAM, 100mbps and 1G on IPv6, and with 2x 960GB Intel SSD only a year old and not much used.

If I cannot figure out the use-case for 2x KS1, I might xfer them to someone here for what I paid for them. What would you do with KS1 if you had a couple? Or I just keep them, what might I do with single disk systems? Please brainstorm me some idea!

What would you do with KS-LE-1?
What would you do with KS1?

What all do people do with their Kimsufi servers?

I'm primarily interested in offsite redundant backup (for cold store type purpose, not needing to access it too much). What are some good storage solutions? Should I consider S3Glacier?Thanks <3

«13

Comments

  • You can't transfer Kimsufi anymore, so forget about it.

  • stonedstoned Member
    edited November 2022

    @dahartigan said:
    You can't transfer Kimsufi anymore, so forget about it.

    Well then aren't I the lucky one. Getting everything I want today! :D

    Please advise if you can about what good use I can put these boxes to then. Thanks.

  • @stoned said:

    Please advise if you can about what good use I can put these boxes to then. Thanks.

    I would suggest idling them.

  • stonedstoned Member
    edited November 2022

    I'll just dump data on them for cold store for redundancy and let them idle. \o/

    Each KS1 can have identical data backed up to it.

    As for KS-LE-1, what would you guys do with yours?

  • Seed.

  • MaouniqueMaounique Host Rep, Veteran

    I am always donating the BW at least for Tor and freenet when I can, especially Tor bridges are needed these days as many dictatorships ban the public Tor IPs. Significant storage allows for back-ups as well.
    Learning is good if you are not an expert already. Play with Wireguard, make all your servers and VMs reside within your "local" network. Harden against attacks. OVH IPv4 ranges are among the most attacked in the world. Make a honeypot.

  • @stoned said:

    I'm primarily interested in offsite redundant backup (for cold store type purpose, not needing to access it too much). What are some good storage solutions? Should I consider S3Glacier?Thanks <3

    Backblaze. S3 compatible bucket storage, available with versioning, works with rclone and every other compatible featureset you can imagine. Absolutely dirt cheap. Combine with a few little changes to your setup on cloudflare and you basically have rolled your own CDN for stupid cheap prices. I highly recommend Backblaze personally, they've been great in my use.

    Thanked by 4stoned amarc abtdw pbx
  • Install Minio on them. Or, Filerun.

  • mgcAnamgcAna Member, Host Rep

    This is also I was thinking to ask. For us there is no use of KS but I see they sell a lot. Other than personal thing, I wonder what uses people are making out of KS.

  • I picked up a KS-1 a couple days ago to migrate my small mastodon instance to. Mastodon likes some space to breath and I ended up with the 2TB HDD.

  • @stoned said: I'll just dump data on them for cold store for redundancy and let them idle.

    exactly. I just use it as storage server for my wordpress blog images.

  • I got 3 KS-1 with 4 TB each. I run glusterfs in dispersed mode with a redundancy parameter similar to RAID5 so basically my gluster cluster can sustain one failure and has ~8 TB of usable storage.
    All three nodes are connected using Wireguard as TLS was way too demanding for these poor souls and it was killing the bandwidth.
    In the end, when using glusterfs native client from another machine connected to all three KS-1, you get almost 200 Mbps of write speed as the native client directly writes to all three units at the same time. The client connects to all three KS-1 using Wireguard as well

  • rm_rm_ IPv6 Advocate, Veteran
    edited November 2022

    Let's list some of the preconceptions preventing you to just use these normally:

    • "but no RAID"; do a daily backup (or more often), which you should anyway, keep an eye on SMART (smartmontools can even send an E-Mail in case there's a concern) and it'll be fine.
    • "but slow CPU"; do not use bloated crap, pick a small and simple CMS, and you can run your website on it, I know I did with dokuwiki.
    • "but only 100 Mbit"; In case your webpage is 100KB on average, enough for 100 concurrent visitors; in case your pages are larger, well then optimize your pages.

    I could run my everything on a KS1, luckily I have a KS4C from a past sale, so using that for now (also single disk).

    Only thing that would strain it, while fitting multiple tasks into a KS1, is a popular seedbox (with qBittorrent), could require some tweaking of priorities and RAM buffers, but still might be good enough for the <100Mbit b/w.

    Thanked by 1stoned
  • MaouniqueMaounique Host Rep, Veteran

    @rm_ said: "but only 100 Mbit"; In case your webpage is 100KB on average, enough for 100 concurrent visitors; in case your pages are larger, well then optimize your pages.

    Not really, only a bot will continuously load pages, most people will load one every few minutes in average.

  • vitalis3vitalis3 Member
    edited November 2022

    Use the servers to power Psychedelics Daily and HashSecurity?

    Thanked by 1dahartigan
  • booby gifs

  • @vitalis3 said:
    Use the servers to power Psychedelics Daily and HashSecurity?

    Hash borg right?

  • dahartigandahartigan Member
    edited November 2022

    @bdl said: booby

    PLS SEND

    Thanked by 1bdl
  • @dahartigan said: Hash borg right?

    Of course.

    Thanked by 1dahartigan
  • @vitalis3 said:

    @dahartigan said: Hash borg right?

    Of course.

    The enlightened one.

  • The cheaper the better

  • @stoned said: As for KS-LE-1, what would you guys do with yours?

    fuck it

  • @Val said:
    I got 3 KS-1 with 4 TB each. I run glusterfs in dispersed mode with a redundancy parameter similar to RAID5 so basically my gluster cluster can sustain one failure and has ~8 TB of usable storage.
    All three nodes are connected using Wireguard as TLS was way too demanding for these poor souls and it was killing the bandwidth.
    In the end, when using glusterfs native client from another machine connected to all three KS-1, you get almost 200 Mbps of write speed as the native client directly writes to all three units at the same time. The client connects to all three KS-1 using Wireguard as well

    What location?

  • @plumberg said:

    @Val said:
    I got 3 KS-1 with 4 TB each. I run glusterfs in dispersed mode with a redundancy parameter similar to RAID5 so basically my gluster cluster can sustain one failure and has ~8 TB of usable storage.
    All three nodes are connected using Wireguard as TLS was way too demanding for these poor souls and it was killing the bandwidth.
    In the end, when using glusterfs native client from another machine connected to all three KS-1, you get almost 200 Mbps of write speed as the native client directly writes to all three units at the same time. The client connects to all three KS-1 using Wireguard as well

    What location?

    1 in RBX, 2 in GRA. It's much harder to grab one from RBX while you regularly see some drops from GRA.
    To get 4 TB, I asked support to replace my disks and got 4 TB the three times. Not sure if I was extremely lucky (especially since it was in distinct DC each time) or if they don't have any 2 TB HDD on hands anymore.

    Thanked by 1plumberg
  • @Val said:

    @plumberg said:

    @Val said:
    I got 3 KS-1 with 4 TB each. I run glusterfs in dispersed mode with a redundancy parameter similar to RAID5 so basically my gluster cluster can sustain one failure and has ~8 TB of usable storage.
    All three nodes are connected using Wireguard as TLS was way too demanding for these poor souls and it was killing the bandwidth.
    In the end, when using glusterfs native client from another machine connected to all three KS-1, you get almost 200 Mbps of write speed as the native client directly writes to all three units at the same time. The client connects to all three KS-1 using Wireguard as well

    What location?

    1 in RBX, 2 in GRA. It's much harder to grab one from RBX while you regularly see some drops from GRA.
    To get 4 TB, I asked support to replace my disks and got 4 TB the three times. Not sure if I was extremely lucky (especially since it was in distinct DC each time) or if they don't have any 2 TB HDD on hands anymore.

    Wow you really got lucky. 4tb is good for cold storage at that price. And the face they offered a replacement.

    What was original disk size? What reason you seeked replacement?

  • @plumberg said:

    @Val said:

    @plumberg said:

    @Val said:
    I got 3 KS-1 with 4 TB each. I run glusterfs in dispersed mode with a redundancy parameter similar to RAID5 so basically my gluster cluster can sustain one failure and has ~8 TB of usable storage.
    All three nodes are connected using Wireguard as TLS was way too demanding for these poor souls and it was killing the bandwidth.
    In the end, when using glusterfs native client from another machine connected to all three KS-1, you get almost 200 Mbps of write speed as the native client directly writes to all three units at the same time. The client connects to all three KS-1 using Wireguard as well

    What location?

    1 in RBX, 2 in GRA. It's much harder to grab one from RBX while you regularly see some drops from GRA.
    To get 4 TB, I asked support to replace my disks and got 4 TB the three times. Not sure if I was extremely lucky (especially since it was in distinct DC each time) or if they don't have any 2 TB HDD on hands anymore.

    Wow you really got lucky. 4tb is good for cold storage at that price. And the face they offered a replacement.

    What was original disk size? What reason you seeked replacement?

    They all were 2 TB disks (so I was already lucky). I seeked replacement because they all were 6 or 7 years old according to smartctl, and it seems OVH support agreed that they needed replacement!

    Thanked by 1plumberg
  • @stoned said: What all do people do with their Kimsufi servers?

    • Backup (2TB for $5.99/mo is pretty decent
    • Security (I'm not running a bank, but if I absolutely need to keep password files or something similar, at least they're on bare metal with dedicated disks instead of a shared KVM host)
    • If RIPE ever asks why I need an ASN and /40 IPv6 allocation in their region instead of my own, at least I can say I have dedicated equipment in 2 OVH datacenters

    @Val said: To get 4 TB, I asked support to replace my disks and got 4 TB the three times.

    Sounds like I need to check the disks in mine. You may be my favorite person of the week, @Val 😃

    Thanked by 1Val
  • Keep me updated on how it goes for you! @jtcx
    Make sure to add in the form the smartctl's output and a little comment like how your server has old disks and you would like newer drives as you'd like to keep your servers for quite some time without issues.
    Also, my KS-1s are on the new OVHCloud manager, I'm not sure if it helped in any ways (do we have better OVH support with the new panel?)

    Thanked by 1vertin
  • @Val said: Make sure to add in the form the smartctl's output and a little comment like how your server has old disks and you would like newer drives

    Did smartctl show any errors, or just age?

  • @jtcx said:

    @Val said: Make sure to add in the form the smartctl's output and a little comment like how your server has old disks and you would like newer drives

    Did smartctl show any errors, or just age?

    Nope, but Powered-on hours amounted 6-7 years!

Sign In or Register to comment.