Howdy, Stranger!

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


Shells Virtual Desktop
BMail.ag - Secure Email Service
Server.net
CPLicense.net
VPS Server
Buy VPN
Vultr
VMs for AI
HostDare
HostDare
ReliableSite White-Label Dedicated Hosting for Resellers
InterServer VPS
BMail.ag - Secure Email Service
Best VPN
High-Performance Bare Metal Server Solutions
Karvl.com
Server Mania Cloud Hosting
DataWagon Hosting
AlphaVPS Hosting
Evoxt.com
Clouvider
VPS Hosting with NVMe
Residential IPs in the US & 4G Mobile Proxies in EU & US with Unlimited Bandwidth
ReliableSite White-Label Dedicated Hosting for Resellers
Rabisu - Hosting Solutions
Shells Virtual Desktop
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.

Pulsed Media 12Th Anniversary Celebration ** FREE Seedboxes AND Special Anniversary Seedboxes!

1151618202126

Comments

  • @elliotc said: My M10G is better than the V10G. I haven't done any bench, just feeling. In fact, there is no need for a bench, because the difference is obvious.

    My M10G is Ryzen 5 5600X, so definitely better than V10G

    Thanked by 1PulsedMedia
  • PulsedMediaPulsedMedia Member, Patron Provider

    @hotsnow said:

    @PulsedMedia said:
    Funny ticket to start the day with, anniversary V10G S on 4xHDD server under load with the YABS FIO results as evidence:

    Yucky Disk Speed. Please check for Chia.

    fio Disk Speed Tests (Mixed R/W 50/50):
    Block Size  4k (IOPS)   64k (IOPS)
    Read    2.49 MB/s (622)     30.27 MB/s (473)
    Write   2.50 MB/s (626)     30.62 MB/s (478)
    Total   4.99 MB/s (1.2k)    60.89 MB/s (951)
            
    Block Size  512k (IOPS)     1m (IOPS)
    ------  --- ----    ---- ----
    Read    119.00 MB/s (232)   145.47 MB/s (142)
    Write   125.32 MB/s (244)   155.16 MB/s (151)
    Total   244.32 MB/s (476)   300.64 MB/s (293) 
    

    Not too shabby for 4x HDD server under load.

    seems this I/O is already very good, much better than my M10G :'(

    Yabs on my M10G

    Tue Feb 15 15:48:50 EET 2022
    
    Basic System Information:
    ---------------------------------
    Processor  : Intel(R) Xeon(R) CPU           X5660  @ 2.80GHz
    CPU cores  : 6 @ 2712.954 MHz
    AES-NI     : ✔ Enabled
    VM-x/AMD-V : ✔ Enabled
    RAM        : 70.8 GiB
    Swap       : 5.6 GiB
    Disk       : 38.0 TiB
    
    fio Disk Speed Tests (Mixed R/W 50/50):
    ---------------------------------
    Block Size | 4k            (IOPS) | 64k           (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 411.00 KB/s    (102) | 5.80 MB/s       (90)
    Write      | 432.00 KB/s    (108) | 6.13 MB/s       (95)
    Total      | 843.00 KB/s    (210) | 11.93 MB/s     (185)
               |                      |                     
    Block Size | 512k          (IOPS) | 1m            (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 22.87 MB/s      (44) | 29.74 MB/s      (29)
    Write      | 24.21 MB/s      (47) | 32.54 MB/s      (31)
    Total      | 47.08 MB/s      (91) | 62.28 MB/s      (60)
    

    That server sure has some I/O load. Try again a bit later, who knows if someone was just doing torrent check(s) at the same time etc.

    That's the thing, single run YABS won't even show the reality, it's just a snapshot at that time, and these servers have hugely varying load.

  • fanfan Veteran

    This M10G raffle is indeed a powerful tiny box:

    Basic System Information:

    Processor : AMD Ryzen 9 3900 12-Core Processor
    CPU cores : 24 @ 2199.357 MHz
    AES-NI : ▒ Enabled
    VM-x/AMD-V : ▒ Enabled
    RAM : 125.9 GiB
    Swap : 29.8 GiB
    Disk : 87.1 TiB

    fio Disk Speed Tests (Mixed R/W 50/50):

    Block Size 4k (IOPS) 64k (IOPS)
    Read 1.93 MB/s (483) 21.53 MB/s (336)
    Write 1.95 MB/s (489) 21.99 MB/s (343)
    Total 3.89 MB/s (972) 43.52 MB/s (679)
    Block Size 512k (IOPS) 1m (IOPS)
    ------ --- ---- ---- ----
    Read 119.62 MB/s (233) 152.93 MB/s (149)
    Write 125.97 MB/s (246) 163.11 MB/s (159)
    Total 245.59 MB/s (479) 316.04 MB/s (308)

    Geekbench 5 Benchmark Test:

    Test | Value
    |
    Single Core | 1298
    Multi Core | 10669
    Full Test | https://browser.geekbench.com/v5/cpu/12794344

    Thanked by 1PulsedMedia
  • @hotsnow said:

    seems this I/O is already very good, much better than my M10G :'(

    Yabs on my M10G

    Tue Feb 15 15:48:50 EET 2022
    
    Basic System Information:
    ---------------------------------
    Processor  : Intel(R) Xeon(R) CPU           X5660  @ 2.80GHz
    CPU cores  : 6 @ 2712.954 MHz
    AES-NI     : ✔ Enabled
    VM-x/AMD-V : ✔ Enabled
    RAM        : 70.8 GiB
    Swap       : 5.6 GiB
    Disk       : 38.0 TiB
    
    fio Disk Speed Tests (Mixed R/W 50/50):
    ---------------------------------
    Block Size | 4k            (IOPS) | 64k           (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 411.00 KB/s    (102) | 5.80 MB/s       (90)
    Write      | 432.00 KB/s    (108) | 6.13 MB/s       (95)
    Total      | 843.00 KB/s    (210) | 11.93 MB/s     (185)
               |                      |                     
    Block Size | 512k          (IOPS) | 1m            (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 22.87 MB/s      (44) | 29.74 MB/s      (29)
    Write      | 24.21 MB/s      (47) | 32.54 MB/s      (31)
    Total      | 47.08 MB/s      (91) | 62.28 MB/s      (60)
    

    yabs is the result at that moment, so it will also change depending on the time of day and the activity of other users.
    (The results on my M10G are good, but this is a coincidence in a sense)

    Tue Feb 15 16:07:07 EET 2022
    
    Block Size | 4k            (IOPS) | 64k           (IOPS)
      ------   | ---            ----  | ----           ----
    Read       | 2.43 MB/s      (609) | 35.55 MB/s     (555)
    Write      | 2.45 MB/s      (613) | 35.90 MB/s     (561)
    Total      | 4.89 MB/s     (1.2k) | 71.46 MB/s    (1.1k)
               |                      |
    Block Size | 512k          (IOPS) | 1m            (IOPS)
      ------   | ---            ----  | ----           ----
    Read       | 144.99 MB/s    (283) | 190.35 MB/s    (185)
    Write      | 152.70 MB/s    (298) | 203.03 MB/s    (198)
    Total      | 297.70 MB/s    (581) | 393.38 MB/s    (383)
    
    ---------------------------------
    
    Tue Feb 15 16:37:41 EET 2022
    
    Block Size | 4k            (IOPS) | 64k           (IOPS)
      ------   | ---            ----  | ----           ----
    Read       | 2.58 MB/s      (645) | 37.99 MB/s     (593)
    Write      | 2.59 MB/s      (647) | 38.26 MB/s     (597)
    Total      | 5.17 MB/s     (1.2k) | 76.26 MB/s    (1.1k)
               |                      |
    Block Size | 512k          (IOPS) | 1m            (IOPS)
      ------   | ---            ----  | ----           ----
    Read       | 123.66 MB/s    (241) | 221.03 MB/s    (215)
    Write      | 130.23 MB/s    (254) | 235.75 MB/s    (230)
    Total      | 253.89 MB/s    (495) | 456.79 MB/s    (445)
    
    Thanked by 1PulsedMedia
  • hotsnowhotsnow Veteran
    edited February 2022

    @PulsedMedia said:
    That server sure has some I/O load. Try again a bit later, who knows if someone was just doing torrent check(s) at the same time etc.

    That's the thing, single run YABS won't even show the reality, it's just a snapshot at that time, and these servers have hugely varying load.

    tested at least five times in recently two days, different time of day, but same results, 40-60MB/s

    seems the M10G with Ryzen are much better

  • @hotsnow said:

    @PulsedMedia said:
    That server sure has some I/O load. Try again a bit later, who knows if someone was just doing torrent check(s) at the same time etc.

    That's the thing, single run YABS won't even show the reality, it's just a snapshot at that time, and these servers have hugely varying load.

    tested at least five times in recently two days, different time of day, but same results, 40-60MB/s

    seems the M10G with Ryzen are much better

    M10G and good neighbors

    CPU: Intel(R) Xeon(R) CPU X5660 @ 2.80GHz
    first day test 146-187 MB/s
    second day test 30-55 MB/s
    right now 15-18 MB/s

    Thanked by 1hotsnow
  • PulsedMediaPulsedMedia Member, Patron Provider

    @tototo said:

    @hotsnow said:

    seems this I/O is already very good, much better than my M10G :'(

    Yabs on my M10G

    Tue Feb 15 15:48:50 EET 2022
    
    Basic System Information:
    ---------------------------------
    Processor  : Intel(R) Xeon(R) CPU           X5660  @ 2.80GHz
    CPU cores  : 6 @ 2712.954 MHz
    AES-NI     : ✔ Enabled
    VM-x/AMD-V : ✔ Enabled
    RAM        : 70.8 GiB
    Swap       : 5.6 GiB
    Disk       : 38.0 TiB
    
    fio Disk Speed Tests (Mixed R/W 50/50):
    ---------------------------------
    Block Size | 4k            (IOPS) | 64k           (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 411.00 KB/s    (102) | 5.80 MB/s       (90)
    Write      | 432.00 KB/s    (108) | 6.13 MB/s       (95)
    Total      | 843.00 KB/s    (210) | 11.93 MB/s     (185)
               |                      |                     
    Block Size | 512k          (IOPS) | 1m            (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 22.87 MB/s      (44) | 29.74 MB/s      (29)
    Write      | 24.21 MB/s      (47) | 32.54 MB/s      (31)
    Total      | 47.08 MB/s      (91) | 62.28 MB/s      (60)
    

    yabs is the result at that moment, so it will also change depending on the time of day and the activity of other users.
    (The results on my M10G are good, but this is a coincidence in a sense)

    Tue Feb 15 16:07:07 EET 2022
    
    Block Size | 4k            (IOPS) | 64k           (IOPS)
      ------   | ---            ----  | ----           ----
    Read       | 2.43 MB/s      (609) | 35.55 MB/s     (555)
    Write      | 2.45 MB/s      (613) | 35.90 MB/s     (561)
    Total      | 4.89 MB/s     (1.2k) | 71.46 MB/s    (1.1k)
               |                      |
    Block Size | 512k          (IOPS) | 1m            (IOPS)
      ------   | ---            ----  | ----           ----
    Read       | 144.99 MB/s    (283) | 190.35 MB/s    (185)
    Write      | 152.70 MB/s    (298) | 203.03 MB/s    (198)
    Total      | 297.70 MB/s    (581) | 393.38 MB/s    (383)
    
    ---------------------------------
    
    Tue Feb 15 16:37:41 EET 2022
    
    Block Size | 4k            (IOPS) | 64k           (IOPS)
      ------   | ---            ----  | ----           ----
    Read       | 2.58 MB/s      (645) | 37.99 MB/s     (593)
    Write      | 2.59 MB/s      (647) | 38.26 MB/s     (597)
    Total      | 5.17 MB/s     (1.2k) | 76.26 MB/s    (1.1k)
               |                      |
    Block Size | 512k          (IOPS) | 1m            (IOPS)
      ------   | ---            ----  | ----           ----
    Read       | 123.66 MB/s    (241) | 221.03 MB/s    (215)
    Write      | 130.23 MB/s    (254) | 235.75 MB/s    (230)
    Total      | 253.89 MB/s    (495) | 456.79 MB/s    (445)
    

    damn, those are some impressive results for in-production 10G server even if i say so myself :)

    @hotsnow said:

    @PulsedMedia said:
    That server sure has some I/O load. Try again a bit later, who knows if someone was just doing torrent check(s) at the same time etc.

    That's the thing, single run YABS won't even show the reality, it's just a snapshot at that time, and these servers have hugely varying load.

    tested at least five times in recently two days, different time of day, but same results, 40-60MB/s

    seems the M10G with Ryzen are much better

    Open a ticket with those results, so we can check if there is an actual issue with the server or just very high load.

    One thing with load is that it by nature goes down over time the longer people have used the server as well. The first 2-3 weeks on any server is brutal, everyone at the same time trying to fill their services, test everything, set things up etc etc. Once things are up and running loads lower as things are how people like then and they already know how things work etc. :)

    Thanked by 1tototo
  • PulsedMediaPulsedMedia Member, Patron Provider

    I wish we could do something about that performance variance and load variance, but not much can be done :( it's very difficult to predict how heavy a new user is, a minimum number of usage needs to be hit per server too etc. :/

    We have a long term plan for that tho, but nothing which can be made to happen in any scale within a year

  • AstroAstro Member
    edited February 2022
    Running fio random mixed R+W disk test with 512k block size..fio Disk Speed Tests (Mixed R/W 50/50):                      ---------------------------------
    Block Size | 4k            (IOPS) | 64k           (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 1.78 MB/s      (445) | 23.52 MB/s     (367)
    Write      | 1.80 MB/s      (451) | 23.94 MB/s     (374)
    Total      | 3.58 MB/s      (896) | 47.47 MB/s     (741)
               |                      |                     
    Block Size | 512k          (IOPS) | 1m            (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 122.85 MB/s    (239) | 139.03 MB/s    (135)
    Write      | 129.37 MB/s    (252) | 148.28 MB/s    (144)
    Total      | 252.22 MB/s    (491) | 287.31 MB/s    (279)
    

    On the Astro deal! Not shabby at all

  • yoursunnyyoursunny Member, IPv6 Advocate

    @PulsedMedia said:
    There's been quite a few now using the script with all the ARR apps, and breaking their things, then expecting us to fix them+compensate / refund.

    A lot of people seem not to realize if they customize their service they are responsible for the customizations / configs. Quite a few refund requests because people broke their lighttpd with custom config.

    Provide a script/button to re-generate everything?

    1. Move all files in the home directory except ~/data and ~/backup.* to ~/backup.$(date +%s).
    2. Download/generate new configs.

    This can instantly undo all customizations and make the account functional again, unless the disk quota is exceeded.

  • PulsedMediaPulsedMedia Member, Patron Provider

    @yoursunny said:

    @PulsedMedia said:
    There's been quite a few now using the script with all the ARR apps, and breaking their things, then expecting us to fix them+compensate / refund.

    A lot of people seem not to realize if they customize their service they are responsible for the customizations / configs. Quite a few refund requests because people broke their lighttpd with custom config.

    Provide a script/button to re-generate everything?

    1. Move all files in the home directory except ~/data and ~/backup.* to ~/backup.$(date +%s).
    2. Download/generate new configs.

    This can instantly undo all customizations and make the account functional again, unless the disk quota is exceeded.

    We already have one, but that needs root access so it's only ever manually used via a ticket. Need to add that button to billing side, same with restart everything, and perhaps start pulling in traffic + disk usage. One thing for a long time we've wanted to do is offer full reprovision button -> remove current and create to new account, tho that needs a once a week limiter or something like that.

  • @Unixfy said:

    @stab said: Is there some way I can fix this myself without having to open a ticket? I tried deleting the files from .bin and .config and the custom lighttpd rules but nothing seems to fix it.

    Usually it's caused by a configuration error, you can try doing this:

    /usr/sbin/lighttpd -f ~/.lighttpd.conf

    Then Lighttpd will throw some (helpful) errors in case your config is wrong :)

    @dgc1980 said:
    touch .lighttpd/custom to make sure the file is there again, and lighttpd should automatically start up again on its own

    Thank you both! I managed to fix it with this info.

  • yoursunnyyoursunny Member, IPv6 Advocate

    @PulsedMedia said:

    @yoursunny said:

    @PulsedMedia said:
    There's been quite a few now using the script with all the ARR apps, and breaking their things, then expecting us to fix them+compensate / refund.

    A lot of people seem not to realize if they customize their service they are responsible for the customizations / configs. Quite a few refund requests because people broke their lighttpd with custom config.

    Provide a script/button to re-generate everything?

    1. Move all files in the home directory except ~/data and ~/backup.* to ~/backup.$(date +%s).
    2. Download/generate new configs.

    This can instantly undo all customizations and make the account functional again, unless the disk quota is exceeded.

    We already have one, but that needs root access so it's only ever manually used via a ticket. Need to add that button to billing side, same with restart everything, and perhaps start pulling in traffic + disk usage. One thing for a long time we've wanted to do is offer full reprovision button -> remove current and create to new account, tho that needs a once a week limiter or something like that.

    Moving and generating files inside the home directory doesn't need root.
    Adding/deleting user account or controlling system-level services needs root, but these parts can be skipped for repairing a user account.

    Full reprovision button would delete all the movies, so it's undesirable in many cases.

    Thanked by 1vimalware
  • PulsedMediaPulsedMedia Member, Patron Provider

    @yoursunny said:

    @PulsedMedia said:

    @yoursunny said:

    @PulsedMedia said:
    There's been quite a few now using the script with all the ARR apps, and breaking their things, then expecting us to fix them+compensate / refund.

    A lot of people seem not to realize if they customize their service they are responsible for the customizations / configs. Quite a few refund requests because people broke their lighttpd with custom config.

    Provide a script/button to re-generate everything?

    1. Move all files in the home directory except ~/data and ~/backup.* to ~/backup.$(date +%s).
    2. Download/generate new configs.

    This can instantly undo all customizations and make the account functional again, unless the disk quota is exceeded.

    We already have one, but that needs root access so it's only ever manually used via a ticket. Need to add that button to billing side, same with restart everything, and perhaps start pulling in traffic + disk usage. One thing for a long time we've wanted to do is offer full reprovision button -> remove current and create to new account, tho that needs a once a week limiter or something like that.

    Moving and generating files inside the home directory doesn't need root.
    Adding/deleting user account or controlling system-level services needs root, but these parts can be skipped for repairing a user account.

    Full reprovision button would delete all the movies, so it's undesirable in many cases.

    Yes that does not need root, but how our account recreation works now also reconfigures everything, and needs to change some nginx etc. settings. so cannot just link to that functionality.

    Some people prefer full reprovision, quite a lot of those requests actually.

  • PulsedMediaPulsedMedia Member, Patron Provider

    Discord WIP - a few people there, and i think we got the roles and channels now roughly OK

    Join us at https://discord.gg/9mAPuRSrpG

    Thanked by 2Astro Ganonk
  • @PulsedMedia said:
    Discord WIP - a few people there, and i think we got the roles and channels now roughly OK

    Join us at https://discord.gg/9mAPuRSrpG

    See you all there

  • I've been testing for a little while. Will definitely renew my M10G 1 TB.

    Setting up Jellyfin was pretty easy, but I realized that they don't have offline download for iOS which is an issue for me. Might need to find a way to make it work with Plex. Emby seems fine, but their paid version is way too expensive.

    Yabs:

    Tue Feb 15 18:19:46 EET 2022

    Basic System Information:

    Processor : AMD Ryzen 5 5600X 6-Core Processor
    CPU cores : 12 @ 2221.747 MHz
    AES-NI : ✔ Enabled
    VM-x/AMD-V : ✔ Enabled
    RAM : 62.9 GiB
    Swap : 59.6 GiB
    Disk : 101.6 TiB

    fio Disk Speed Tests (Mixed R/W 50/50):

    Block Size 4k (IOPS) 64k (IOPS)
    Read 2.44 MB/s (611) 22.47 MB/s (351)
    Write 2.46 MB/s (616) 22.94 MB/s (358)
    Total 4.91 MB/s (1.2k) 45.42 MB/s (709)
    Block Size 512k (IOPS) 1m (IOPS)
    ------ --- ---- ---- ----
    Read 80.53 MB/s (157) 102.55 MB/s (100)
    Write 84.81 MB/s (165) 109.37 MB/s (106)
    Total 165.35 MB/s (322) 211.92 MB/s (206)

    iperf3 Network Speed Tests (IPv4):

    Provider | Location (Link) | Send Speed | Recv Speed
    | | |
    Clouvider | London, UK (10G) | 2.95 Gbits/sec | 2.45 Gbits/sec
    Online.net | Paris, FR (10G) | 7.04 Gbits/sec | 5.89 Gbits/sec
    WorldStream | The Netherlands (10G) | busy | 721 Mbits/sec
    WebHorizon | Singapore (400M) | 391 Mbits/sec | 503 Mbits/sec
    Clouvider | NYC, NY, US (10G) | 1.96 Gbits/sec | 628 Mbits/sec
    Velocity Online | Tallahassee, FL, US (10G) | 3.19 Gbits/sec | 3.64 Gbits/sec
    Clouvider | Los Angeles, CA, US (10G) | 1.13 Gbits/sec | 311 Mbits/sec
    Iveloz Telecom | Sao Paulo, BR (2G) | 768 Mbits/sec | 236 Mbits/sec

    Geekbench 5 Benchmark Test:

    Test | Value
    |
    Single Core | 1644
    Multi Core | 6828
    Full Test | https://browser.geekbench.com/v5/cpu/12796267

    Thanked by 1PulsedMedia
  • PulsedMediaPulsedMedia Member, Patron Provider

    @Wicked said:
    I've been testing for a little while. Will definitely renew my M10G 1 TB.

    Setting up Jellyfin was pretty easy, but I realized that they don't have offline download for iOS which is an issue for me. Might need to find a way to make it work with Plex. Emby seems fine, but their paid version is way too expensive.

    Yabs:

    Tue Feb 15 18:19:46 EET 2022

    Basic System Information:

    Processor : AMD Ryzen 5 5600X 6-Core Processor
    CPU cores : 12 @ 2221.747 MHz
    AES-NI : ✔ Enabled
    VM-x/AMD-V : ✔ Enabled
    RAM : 62.9 GiB
    Swap : 59.6 GiB
    Disk : 101.6 TiB

    fio Disk Speed Tests (Mixed R/W 50/50):

    Block Size 4k (IOPS) 64k (IOPS)
    Read 2.44 MB/s (611) 22.47 MB/s (351)
    Write 2.46 MB/s (616) 22.94 MB/s (358)
    Total 4.91 MB/s (1.2k) 45.42 MB/s (709)
    Block Size 512k (IOPS) 1m (IOPS)
    ------ --- ---- ---- ----
    Read 80.53 MB/s (157) 102.55 MB/s (100)
    Write 84.81 MB/s (165) 109.37 MB/s (106)
    Total 165.35 MB/s (322) 211.92 MB/s (206)

    iperf3 Network Speed Tests (IPv4):

    Provider | Location (Link) | Send Speed | Recv Speed
    | | |
    Clouvider | London, UK (10G) | 2.95 Gbits/sec | 2.45 Gbits/sec
    Online.net | Paris, FR (10G) | 7.04 Gbits/sec | 5.89 Gbits/sec
    WorldStream | The Netherlands (10G) | busy | 721 Mbits/sec
    WebHorizon | Singapore (400M) | 391 Mbits/sec | 503 Mbits/sec
    Clouvider | NYC, NY, US (10G) | 1.96 Gbits/sec | 628 Mbits/sec
    Velocity Online | Tallahassee, FL, US (10G) | 3.19 Gbits/sec | 3.64 Gbits/sec
    Clouvider | Los Angeles, CA, US (10G) | 1.13 Gbits/sec | 311 Mbits/sec
    Iveloz Telecom | Sao Paulo, BR (2G) | 768 Mbits/sec | 236 Mbits/sec

    Geekbench 5 Benchmark Test:

    Test | Value
    |
    Single Core | 1644
    Multi Core | 6828
    Full Test | https://browser.geekbench.com/v5/cpu/12796267

    Plex is quite easy to setup too, if you don't mind all the privacy and data safety issues with Plex.

  • @PulsedMedia doesnt it need a dedicated IP?

    @Wicked im currently mounting the seedbox remotely on a VPS which has plex installed. Works like a charm.

  • @Astro said:
    @PulsedMedia doesnt it need a dedicated IP?

    @Wicked im currently mounting the seedbox remotely on a VPS which has plex installed. Works like a charm.

    mounting the seedbox remotely on a VPS

    Thinking about that too, how many server resource(ram, cpu juice) for your plex instance?

    Thanked by 1vimalware
  • PulsedMediaPulsedMedia Member, Patron Provider

    @Astro said:
    @PulsedMedia doesnt it need a dedicated IP?

    @Wicked im currently mounting the seedbox remotely on a VPS which has plex installed. Works like a charm.

    Probably yes, so should only be used on a MDS

  • Local download speed is a bit of hit or miss for me and the pattern is quite random.

    Getting ~25MB/s if I am lucky and most of time only 500KB/s. Showing the same on HTTPS download, multiple FTP clients, rsync + ssh, sshfs + rsync, wget etc. Also showing the same on both my Kimsufi/SYS dedis in BHS.

    Overall I am getting 2-5MB/s in average, not sure where the bottleneck is, will test from my online dedi later.

    Any suggestions would be great.

  • PulsedMediaPulsedMedia Member, Patron Provider

    @david_W said:
    Local download speed is a bit of hit or miss for me and the pattern is quite random.

    Getting ~25MB/s if I am lucky and most of time only 500KB/s. Showing the same on HTTPS download, multiple FTP clients, rsync + ssh, sshfs + rsync, wget etc. Also showing the same on both my Kimsufi/SYS dedis in BHS.

    Overall I am getting 2-5MB/s in average, not sure where the bottleneck is, will test from my online dedi later.

    Any suggestions would be great.

    MTR both directions on all conditions is the only way to know what's going on. Internet is big and complicated

  • yoursunnyyoursunny Member, IPv6 Advocate
    edited February 2022

    @david_W said:
    Local download speed is a bit of hit or miss for me and the pattern is quite random.

    Getting ~25MB/s if I am lucky and most of time only 500KB/s. Showing the same on HTTPS download, multiple FTP clients, rsync + ssh, sshfs + rsync, wget etc. Also showing the same on both my Kimsufi/SYS dedis in BHS.

    Overall I am getting 2-5MB/s in average, not sure where the bottleneck is, will test from my online dedi later.

    Any suggestions would be great.

    Despite all the routing adjustments, the lack of BBR congestion control is causing performance bottleneck on TCP traffic.
    This week I added an HTTP reverse proxy on a server in France (Evolution Host in OVH Roubaix), then:

    • The proxy pulls from PulsedMedia over TCP CUBIC. This flow runs within Europe and has acceptable performance.
    • My tablet pulls from the proxy over TCP BBR. This flow runs across the Atlantic where BBR improves performance.
    • The proxy itself listens on wg0 interface so that only my tablet can connect to it.

    When I had end-to-end TCP CUBIC (directly from USA to Finland, or indirectly going through WireGuard VPN in France), movies over 2GB/hour are unusable except early morning.
    Now I have TCP BBR for the most distance, I can stream 2GB/hour movies without pausing for buffering, including in the evening congestion hours.

    Thanked by 2elliotc vimalware
  • @yoursunny said:

    @david_W said:
    Local download speed is a bit of hit or miss for me and the pattern is quite random.

    Getting ~25MB/s if I am lucky and most of time only 500KB/s. Showing the same on HTTPS download, multiple FTP clients, rsync + ssh, sshfs + rsync, wget etc. Also showing the same on both my Kimsufi/SYS dedis in BHS.

    Overall I am getting 2-5MB/s in average, not sure where the bottleneck is, will test from my online dedi later.

    Any suggestions would be great.

    Despite all the routing adjustments, the lack of BBR congestion control is causing performance bottleneck on TCP traffic.
    This week I added an HTTP reverse proxy on a server in France (Evolution Host in OVH Roubaix), then:

    • The proxy pulls from PulsedMedia over TCP CUBIC. This flow runs within Europe and has acceptable performance.
    • My tablet pulls from the proxy over TCP BBR. This flow runs across the Atlantic where BBR improves performance.
    • The proxy itself listens on wg0 interface so that only my tablet can connect to it.

    When I had end-to-end TCP CUBIC (directly from USA to Finland, or indirectly going through WireGuard VPN in France), movies over 2GB/hour are unusable except early morning.
    Now I have TCP BBR for the most distance, I can stream 2GB/hour movies without pausing for buffering, including in the evening congestion hours.

    Can I simplify the above steps to use udp base vpn in Europe instand of a proxy?

  • yoursunnyyoursunny Member, IPv6 Advocate

    @elliotc said:

    @yoursunny said:

    @david_W said:
    Local download speed is a bit of hit or miss for me and the pattern is quite random.

    Getting ~25MB/s if I am lucky and most of time only 500KB/s. Showing the same on HTTPS download, multiple FTP clients, rsync + ssh, sshfs + rsync, wget etc. Also showing the same on both my Kimsufi/SYS dedis in BHS.

    Overall I am getting 2-5MB/s in average, not sure where the bottleneck is, will test from my online dedi later.

    Any suggestions would be great.

    Despite all the routing adjustments, the lack of BBR congestion control is causing performance bottleneck on TCP traffic.
    This week I added an HTTP reverse proxy on a server in France (Evolution Host in OVH Roubaix), then:

    • The proxy pulls from PulsedMedia over TCP CUBIC. This flow runs within Europe and has acceptable performance.
    • My tablet pulls from the proxy over TCP BBR. This flow runs across the Atlantic where BBR improves performance.
    • The proxy itself listens on wg0 interface so that only my tablet can connect to it.

    When I had end-to-end TCP CUBIC (directly from USA to Finland, or indirectly going through WireGuard VPN in France), movies over 2GB/hour are unusable except early morning.
    Now I have TCP BBR for the most distance, I can stream 2GB/hour movies without pausing for buffering, including in the evening congestion hours.

    Can I simplify the above steps to use udp base vpn in Europe instand of a proxy?

    As I stated above, no!
    I already tried going through WireGuard VPN in France.
    It may get around congestion, but you are still running TCP CUBIC end-to-end which suffers from the effects of bufferbloat.

    Thanked by 1vimalware
  • kyroskohkyroskoh Member
    edited February 2022

    @yoursunny said:

    @david_W said:
    Local download speed is a bit of hit or miss for me and the pattern is quite random.

    Getting ~25MB/s if I am lucky and most of time only 500KB/s. Showing the same on HTTPS download, multiple FTP clients, rsync + ssh, sshfs + rsync, wget etc. Also showing the same on both my Kimsufi/SYS dedis in BHS.

    Overall I am getting 2-5MB/s in average, not sure where the bottleneck is, will test from my online dedi later.

    Any suggestions would be great.

    Despite all the routing adjustments, the lack of BBR congestion control is causing performance bottleneck on TCP traffic.
    This week I added an HTTP reverse proxy on a server in France (Evolution Host in OVH Roubaix), then:

    • The proxy pulls from PulsedMedia over TCP CUBIC. This flow runs within Europe and has acceptable performance.
    • My tablet pulls from the proxy over TCP BBR. This flow runs across the Atlantic where BBR improves performance.
    • The proxy itself listens on wg0 interface so that only my tablet can connect to it.

    When I had end-to-end TCP CUBIC (directly from USA to Finland, or indirectly going through WireGuard VPN in France), movies over 2GB/hour are unusable except early morning.
    Now I have TCP BBR for the most distance, I can stream 2GB/hour movies without pausing for buffering, including in the evening congestion hours.

    TCP BBR ftw!~ benefits of BBR = long haul links (those who aren't from EU regions). Due to the fact that BBR ramps up to the optimal sending rate aggressively, causing your video stream to load even faster. Downfall for its tendency to consume all available bandwidth and pushing out other TCP streams that use say Cubic or different congestion algorithms. BBRv2 is supposed to resolve some of these challenges.

    @yoursunny Are u using BBRv2?

    Thanked by 1yoursunny
  • One Astro plan back in stock! Go get it!

  • M10G more!!!!!

  • sezingsezing Member, Host Rep

    @Astro said:
    One Astro plan back in stock! Go get it!

    I got it.

    Is price recurring, right?

Sign In or Register to comment.