Howdy, Stranger!

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


Beta testers needed for cloud project - Page 2
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.

Beta testers needed for cloud project

2»

Comments

  • I'm interested. I will do what I always do, that is to detect (passively) and report malware-infected computers so that their (victim) owners get a chance to fix the problem. HAProxy is my detection tool, which acts as a honeypot with no interaction, waiting patiently for malicious connection attempts.

  • RickBakkrRickBakkr Member, Patron Provider, LIR
    edited December 2020

    This seems a great way to create an environment to build (and play around with) a Kubernetes cluster. Friends of mine have been telling me I should and how wonderful it is, but never saw a great environment to spin the dynamic nodes for it.

    Also, having high performant infra of some kind at the 'neighbours' in Belgium has always been a secret pleasure for me. Would be lovely if you were to let me in.

    Bedankt kerel, or in case you're Walloon: merci beaucoup!

    (Just noticed the "No Providers" rule; does being a LIR around here, make me rule out?)

    Thanked by 11nf
  • jsgjsg Member, Resident Benchmarker

    I can offer what you require and I'd focus on benchmarking, testing and playing with multiple (less common) linux VPS and BSD. Am also interested in non-Ryzen.
    Bonus: you'd get a qualified review; could do/am willing to do one intermediate and one final at the end.

    Thanked by 1lokuzard
  • MannDudeMannDude Host Rep, Veteran
    edited December 2020

    I'm game. I'm poor nowadays and wouldn't mind a system to play with to get reintroduced back into the modern market. Been using VPSes for going on about ten years now and it's always a pleasure to see how the industry progresses over time. Glad it's not just another random OpenVZ/SolusVM offer on some rented dedis like what seems to dominate a lot of these spaces...

    I'll put it to good use.

    Thanked by 1lokuzard
  • interested.
    any slot for me?

    i'm using it for webhosting.

  • Currently using openVZ for my gitlab instance and thinking of moving away from it for performance reasons and love for ryzen.

    so i can install gitlab in this, if you still need a beta tester

  • Well, the response has been overwhelming. I ended up adding 13 people so 3 more then I had in mind.

    Oh well it's Sinterklaas today ;)

    I will also try and monitor activity a bit, if someone has not been using their service within a week or so I will pick someone else.

    Thanks all!!

  • jsgjsg Member, Resident Benchmarker

    Intermediate benchmark/review:

    Machine: amd64, Arch.: amd64, Model: Common KVM processor
    OS, version: FreeBSD 12.2, Mem.: 1.986 GB
    CPU - Cores: 2, Family/Model/Stepping: 15/6/1
    Cache: 64K/64K L1d/L1i, 512K L2, 16M L3
    Std. Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat
              pse36 cflsh mmx fxsr sse sse2 htt sse3 cx16 **aes** hypervisor
    Ext. Flags: syscall nx lm lahf_lm cmp_legacy
    
    ProcMem SC: avg 279.5 - min 272.1, max 284.3
    ProcMem MC: avg 551.8 - min 532.6, max 570.0
    

    Doesn't feel like an Epyc, more like a 26xx v2 (which still is not at all bad). AES and Hypervisor flags but no popcnt. Seems not overbooked, pretty much like a VDS, nice.

    Write seq.: avg 570.05 - min 489.54 (85.9%), max 631.17 (110.7%)
    Write rnd.: avg 4344.66 - min 3725.02 (85.7%), max 4884.05 (112.4%)
    Read seq.:  avg 315.15 - min 267.92 (85.0%), max 345.83 (109.7%)
    Read rnd.:  avg 3529.06 - min 2544.56 (72.1%), max 4807.78 (136.2%)
    --- Disk - Sync/Direct ---
    Write seq.: avg 8.51 - min 8.00 (94.0%), max 8.69 (102.1%)
    Write rnd.: avg 17.38 - min 16.83 (96.8%), max 18.15 (104.4%)
    Read seq.:  avg 972.10 - min 792.94 (81.6%), max 1434.00 (147.5%)
    Read rnd.:  avg 142.16 - min 131.91 (92.8%), max 154.45 (108.6%)
    

    Either a SSD or a not really fast NVMe. Largely tight spread though. All in all (partly upper) mid range I'd say.

    US LAX lax.download.datapacket.com
      DL [MB/s]: avg 37.81 - min 6.07 (16.1%), max 45.11 (119.3%)
      Ping:      avg 149.9 - min 144.8 (96.6%), max 155.3 (103.6%)
      Web ping:  avg 151.2 - min 145.0 (95.9%), max 155.3 (102.7%)
    
    NO OSL speedtest.osl01.softlayer.com
      DL [MB/s]: avg 182.20 - min 116.96 (64.2%), max 202.93 (111.4%)
      Ping:      avg 29.1 - min 28.6 (98.3%), max 38.8 (133.3%)
      Web ping:  avg 29.2 - min 28.7 (98.1%), max 38.8 (132.7%)
    
    US SJC speedtest.sjc01.softlayer.com
      DL [MB/s]: avg 38.74 - min 25.51 (65.8%), max 43.00 (111.0%)
      Ping:      avg 146.7 - min 145.8 (99.4%), max 147.3 (100.4%)
      Web ping:  avg 332.8 - min 145.9 (43.8%), max 2235.2 (671.5%)
    
    IT MIL speedtest.mil01.softlayer.com
      DL [MB/s]: avg 190.35 - min 14.35 (7.5%), max 219.21 (115.2%)
      Ping:      avg 24.9 - min 24.5 (98.5%), max 33.4 (134.2%)
      Web ping:  avg 50.6 - min 24.5 (48.4%), max 843.4 (1667.3%)
    
    FR PAR speedtest.par01.softlayer.com
      DL [MB/s]: avg 362.42 - min 132.29 (36.5%), max 405.82 (112.0%)
      Ping:      avg 13.8 - min 13.5 (97.6%), max 16.3 (117.8%)
      Web ping:  avg 44.2 - min 13.6 (30.8%), max 726.3 (1643.3%)
    
    SG SGP mirror.sg.leaseweb.net
      DL [MB/s]: avg 34.87 - min 1.98 (5.7%), max 40.88 (117.2%)
      Ping:      avg 158.4 - min 158.3 (99.9%), max 159.3 (100.5%)
      Web ping:  avg 158.5 - min 158.3 (99.9%), max 159.4 (100.5%)
    
    GR UNK speedtest.ftp.otenet.gr
      DL [MB/s]: avg 111.44 - min 78.22 (70.2%), max 124.19 (111.4%)
      Ping:      avg 78596.9 - min 50.5 (0.1%), max 801840.3 (1020.2%)
      Web ping:  avg 78641.0 - min 50.5 (0.1%), max 801840.3 (1019.6%)
    
    US WDC mirror.wdc1.us.leaseweb.net
      DL [MB/s]: avg 59.93 - min 1.93 (3.2%), max 73.83 (123.2%)
      Ping:      avg 91.2 - min 85.6 (93.8%), max 95.8 (105.0%)
      Web ping:  avg 91.6 - min 85.7 (93.5%), max 96.1 (104.9%)
    
    DE FRA speedtest.fra02.softlayer.com
      DL [MB/s]: avg 304.82 - min 78.26 (25.7%), max 372.58 (122.2%)
      Ping:      avg 15.4 - min 15.1 (97.8%), max 18.9 (122.4%)
      Web ping:  avg 22.6 - min 15.1 (66.9%), max 242.1 (1072.3%)
    
    RU MOS speedtest.hostkey.ru
      DL [MB/s]: avg 111.08 - min 45.55 (41.0%), max 126.67 (114.0%)
      Ping:      avg 46.4 - min 46.2 (99.6%), max 47.1 (101.5%)
      Web ping:  avg 48.0 - min 46.4 (96.7%), max 52.9 (110.2%)
    
    US DAL speedtest.dal05.softlayer.com
      DL [MB/s]: avg 48.74 - min 29.92 (61.4%), max 53.86 (110.5%)
      Ping:      avg 112.1 - min 111.0 (99.0%), max 118.8 (106.0%)
      Web ping:  avg 116.2 - min 111.1 (95.6%), max 183.4 (157.8%)
    
    UK LON speedtest.lon02.softlayer.com
      DL [MB/s]: avg 501.66 - min 20.78 (4.1%), max 627.40 (125.1%)
      Ping:      avg 9.1 - min 8.7 (95.8%), max 18.2 (200.3%)
      Web ping:  avg 9.4 - min 8.8 (93.9%), max 18.2 (194.3%)
    
    US NYC nyc.download.datapacket.com
      DL [MB/s]: avg 76.87 - min 30.38 (39.5%), max 89.90 (117.0%)
      Ping:      avg 75.8 - min 73.0 (96.3%), max 80.6 (106.4%)
      Web ping:  avg 76.7 - min 73.1 (95.4%), max 81.0 (105.7%)
    
    RO BUC 185.183.99.8
      DL [MB/s]: avg 125.88 - min 75.60 (60.1%), max 139.07 (110.5%)
      Ping:      avg 41.0 - min 40.9 (99.8%), max 41.4 (101.0%)
      Web ping:  avg 41.1 - min 40.9 (99.5%), max 41.5 (101.0%)
    
    FR_RB  lg-fr.hostnamaste.com
      DL [MB/s]: avg 345.69 - min 202.16 (58.5%), max 440.89 (127.5%)
      Ping:      avg 12.5 - min 12.1 (96.9%), max 20.9 (167.4%)
      Web ping:  avg 12.6 - min 12.1 (95.9%), max 20.9 (165.7%)
    

    Similar to disk. Largely (partly upper) mid range. The fastest, UK_LON is about 500 Mb/s, the other big EU 3 are in the +-300 Mb/s range. No really bad result, largely stable, speed to the Americas and to Asia OK.

    All in all nothing to write home about but not at all bad. For the intended price though ... hmmm ...

    Thanked by 1marvel
  • Count me in
    @marvel

  • Count me in
    @marvel

  • I am in

  • fadedmaplefadedmaple Member
    edited January 2021

    I love testing and I haven't used a Belgian vps, But I don't have 50 posts and thanks ...

  • @fadedmaple said:
    I love testing and I haven't used a Belgian vps, But I don't have 50 posts and thanks ...

    @marvel still has a 30% off for you :D

    https://www.lowendtalk.com/discussion/comment/3187313/#Comment_3187313

  • @jsg said:
    All in all nothing to write home about but not at all bad. For the intended price though ... hmmm ...

    Thanks for testing this! First time I even see such an extensive performance test. Great to see the numbers.

    Keep in mind though it's all central storage. You can't expect local disk performance but you get HA, live migration etc. in return.

    How much customers would really need the speed of a local NVMe drive anyway. Perhaps 0.01%. I would rather want stability and longer uptimes, that's where I'm going for and yes in return you will get intermediate performance but still good enough for the majority of the people plus you have less limitations like local available disk space.

    The EPYCs are indeed Zen 1 so a bit more slow in numbers. The Ryzens are much faster but the EPYCs come with more threads, so I will sell both later on and customers can decide. I would assume EPYC is more suitable for e.g. web servers and I would be able to sell those at a cheaper rate. Also, the Ryzens have no dual PSU and no LACP on the public net so when a node fails HA has to boot it up on another one. The EPYCs have all double hardware.

    As for this module, this beta is over. This module is just too buggy. I'm currently messing around with Openstack and Fleio which looks much better already but is very complex to set up. So in the meantime I will just sell preconfigured machines to keep my resources from running idle :smile:

    And about the price, yes I realize it's more expensive but Belgium is a niche location. Power and bandwidth is really expensive here, so I have to make my setup stand out and provide better service to compete with the others.

    Thanks again!

  • I quite liked the service, very flexible and the firewall was a treat. @marvel will eventually implement it another way.

    He has moved me to a ryzen, I think you might find you will lose it on a reboot. But have a ryzen there instead.

    Thanked by 1marvel
  • jsgjsg Member, Resident Benchmarker
    edited January 2021

    @marvel

    My review wasn't a complaint but simply a stating of facts, observations and impressions.

    I personally am not disk-crazy anyway; I still happily run a couple of spinning rust driven VPSs and would look for NVMe only for specific use cases. Btw. your disk performance numbers are not bad (but neither mouth watering) and I follow your argumentation, but I also see that a very considerable part of VPS buyers are numbers-porn driven. Similarly I'm personally not too concerned about processor performance neither, anything north of 200 (in my benchmark) is good enough, anything north of 300 is really great, and Ryzen 3|59xx speed is insane, but again, many, many potential buyers think their website will be slow with anything other than the most current Ryzen and NVMe.

    Where I do see a problem though is with your prices because AMS, PAR, and FRA are "just around the corner" and I guess very few will see having their stuff in Belgium as particularly desirable and pay extra for that.
    My advice would be to follow the route you've already started to walk, the route of redundancy, high availability, etc because that actually is something people (albeit to a lesser degree here at LET I guess) are willing to pay a higher price for.

    Thanked by 1ferri
  • @jsg said:
    @marvel

    My review wasn't a complaint but simply a stating of facts, observations and impressions.

    I personally am not disk-crazy anyway; I still happily run a couple of spinning rust driven VPSs and would look for NVMe only for specific use cases. Btw. your disk performance numbers are not bad (but neither mouth watering) and I follow your argumentation, but I also see that a very considerable part of VPS buyers are numbers-porn driven. Similarly I'm personally not too concerned about processor performance neither, anything north of 200 (in my benchmark) is good enough, anything north of 300 is really great, and Ryzen 3|59xx speed is insane, but again, many, many potential buyers think their website will be slow with anything other than the most current Ryzen and NVMe.

    Where I do see a problem though is with your prices because AMS, PAR, and FRA are "just around the corner" and I guess very few will see having their stuff in Belgium as particularly desirable and pay extra for that.
    My advice would be to follow the route you've already started to walk, the route of redundancy, high availability, etc because that actually is something people (albeit to a lesser degree here at LET I guess) are willing to pay a higher price for.

    You could also say I'm right in the middle of AMS, FRA, PAR and LON ;)

    We'll see, AMS is like 150k so I can always pick everything up move it over there in a weekend in case this fails :blush:

Sign In or Register to comment.