Howdy, Stranger!

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


Budget VPS Comparison
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.

Budget VPS Comparison

Hi everyone!

This is my first post here, and I wanted to share a blog post I just published after testing a bunch of VPS plans this month. I found many on LEB and had a lot of fun benchmarking. Happy to take comments and criticisms!

https://workingconcept.com/blog/2018-budget-vps-survey

«1

Comments

  • bookmarked!

    Thanked by 1matts
  • Good reviews!!!

    Thanked by 1matts
  • well done buddy.

    Thanked by 1matts
  • noted.
    thank you

    Thanked by 1matts
  • Thanks for the kind words. :)

  • Nice work, but if you call Ramnode, Digital Ocean, Linode and Vulture "budget" VPSes you may be in the wrong place.

    Thanked by 1matts
  • mattsmatts Member
    edited October 2018

    Thanks @donli!

    I'd still argue that RamNode offers budget VPSes, or at least skirts the edge of the category. As I wrote though, I added Digital Ocean, Linode, Vultr, and AWS as "established providers" for the sake of interesting comparison.

  • I have heard SSDNode a lot but it kind of looks like a trap to me and I hesitate to give it a shot. Hope you could keep monitoring the performance and update your post a few months later.

    Thanked by 1matts
  • mattsmatts Member
    edited October 2018

    @Kiwi83 said:
    I have heard SSDNode a lot but it kind of looks like a trap to me and I hesitate to give it a shot. Hope you could keep monitoring the performance and update your post a few months later.

    I'm in for a year at this point, and I'd be happy to share an update later. I'm still just as skeptical, and yet aside from slow-ish support tickets the service has been fine.

  • Prepaying for 3 years to get huge discount just sounds like how some crap providers lure people to use their service. But deep down in my heart I always hope SSDNode is the unicorn I am always looking for.

  • @Kiwi83 said:
    Prepaying for 3 years to get huge discount just sounds like how some crap providers lure people to use their service. But deep down in my heart I always hope SSDNode is the unicorn I am always looking for.

    I'm more put out by the false urgency of their sale counter, which I've seen reset several times without impacting the "sale." I also won't pay for more than a year at a time for anything other than domain registrations.

    I too am a unicorn hunter, though. :)

    Thanked by 1vimalware
  • WOW.
    what a nice entry here.
    Welcome, great job!

    Thanked by 2matts Ympker
  • Hi everybody! It's been a year so I thought I'd post an update: https://workingconcept.com/blog/2019-vps-provider-update

    Just some notes and a few more benchmarks.

  • Wow a host that exposes 16 cores to a small VPS. That's new to me.

  • @CyberneticTitan said:
    Wow a host that exposes 16 cores to a small VPS. That's new to me.

    Me too. Seems pretty strange, but I couldn't resist giving it a try.

  • @matts said:

    @CyberneticTitan said:
    Wow a host that exposes 16 cores to a small VPS. That's new to me.

    Me too. Seems pretty strange, but I couldn't resist giving it a try.

    Some Korean companies does something similar. All VMs in the same node share the whole CPU. Although they don't work like your benchmarks since those are oversold as hell

    Thanked by 1matts
  • Good updates, and nice benchmarks.

    I guess you are going to do it once a year. So I have just bookmarked your site. :)

    Thanked by 1matts
  • I use HyperExpert and really like it.

    Thanked by 2matts hyperexpert
  • HostUpHostUp Member, Host Rep

    @matts said:
    Hi everybody! It's been a year so I thought I'd post an update: https://workingconcept.com/blog/2019-vps-provider-update

    Just some notes and a few more benchmarks.

    Pretty cool! I really think you should include vultr's high compute servers as well.

    Thanked by 1matts
  • @matts said:
    Hi everybody! It's been a year so I thought I'd post an update: https://workingconcept.com/blog/2019-vps-provider-update

    Just some notes and a few more benchmarks.

    The information is very well collated and structured. You gave me some ideas! Thanks for the posts...

    Thanked by 1matts
  • jsgjsg Member, Resident Benchmarker

    Nice overview review and nicely presented too, kudos.

    Hint: try to always find and include the CPU version, too (e.g. 2670 v3). Looking at that might also provide some clue wrt your question.

    Thanked by 1matts
  • You have given me some ideas how to create a VPS index. Let me create my formula and then maybe I can create a simple chart comparing BF offers.

    Thanked by 1matts
  • hiphiphip0hiphiphip0 Member
    edited October 2019

    @matts
    I learned individual core speed is critical for PHP (which makes sense), and that MySQL >can scream with multiple cores.

    Then put PHP on a faster core VPS and put MYSQL on a multi-core VPS.
    I personally feel that for PHP, E3 is faster than E5, and ultra-fast IO will make the website feel smooth, like the previous bandwagonhost ones (now the IO is normal).

    Thanked by 1matts
  • @hiphiphip0 said:

    @matts
    I learned individual core speed is critical for PHP (which makes sense), and that MySQL >can scream with multiple cores.

    Then put PHP on a faster core VPS and put MYSQL on a multi-core VPS.
    I personally feel that for PHP, E3 is faster than E5, and ultra-fast IO will make the website feel smooth, like the previous bandwagonhost ones (now the IO is normal).

    I don't usually split projects onto multiple servers—at least not my own—but that's a great idea! I've also felt similar about E3 vs. E5 despite what I've read. In the beginning I set up a CMS project on each server to see what its render times looked like and how it felt. Multiple cores, high IO scores and low latency (distance to datacenter) were always things I could vaguely feel.

    But I started to streamline my testing process because I should be doing other things. :)

  • @HostUp said:

    @matts said:
    Hi everybody! It's been a year so I thought I'd post an update: https://workingconcept.com/blog/2019-vps-provider-update

    Just some notes and a few more benchmarks.

    Pretty cool! I really think you should include vultr's high compute servers as well.

    Oh you're right, I'm not sure how I missed Vultr's High Frequency NVMe offering. Thank you.

  • It would also help to add more providers and make it a more comprehensive list.

    Thanked by 1matts
  • @WiredBlade said:
    It would also help to add more providers and make it a more comprehensive list.

    I'd love to do that if I could justify the amount of time it'd take. I'm just sharing what I've gleaned in my spare time, but IMO tools like serverscope.io are more useful for making more broad comparisons. (I also miss ServerBear.)

  • @matts said:
    I don't usually split projects onto multiple servers—at least not my own—but that's a great idea! I've also felt similar about E3 vs. E5 despite what I've read. In the beginning I set up a CMS project on each server to see what its render times looked like and how it felt. Multiple cores, high IO scores and low latency (distance to datacenter) were always things I could vaguely feel.

    But I started to streamline my testing process because I should be doing other things. :)

    @matts said:
    Oh you're right, I'm not sure how I missed Vultr's High Frequency NVMe offering. Thank you.

    Hi five! That's me. It's like old times when you hear noise from disk, you know some software is eating cpu. I always put web and data on same server too, but for better restore, maybe I would put all databases on one server with multi backup methods.
    I'm moving sites to vultr high frequence VM too, for performance and sleep tight.
    I use runcloud and self made scripts to manage server now, centminmod's document is too detailed, everytime I tried to use it for websites, all end up with reading it's forum posts.
    For backup, I'm using automysqlbackup (github fork also no update now) introduced by Tim@hostigation and rsnapshot (simpler than duplicity).
    Does restic support pulling backup from backup server?
    If the backup is pushing from web server -> backup server, then if web server is compromised, hacker would gain backup server access and destroy the backups too.

    Thanked by 1matts
  • @hiphiphip0 said:
    Does restic support pulling backup from backup server?
    If the backup is pushing from web server -> backup server, then if web server is compromised, hacker would gain backup server access and destroy the backups too.

    Restic makes it easy to mount a volume from the backup server, and regardless of whether it's read-only it includes tools for pruning backups—so I guess a compromised server could mean compromised backups. I always turn on B2 and S3 lifecycles though, which only act as soft-deletes so I'd still be able to retrieve backups for a certain amount of time. I suppose you could also mirror the backup server/volume separately if that was a concern.

  • Great post!

    Want to see indepth comparisons between VPS panel too

    Thanked by 1matts
Sign In or Register to comment.