Howdy, Stranger!

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


Why Google services (e.g blogger ) faster than Hetzner - 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.

Why Google services (e.g blogger ) faster than Hetzner

2»

Comments

  • JohnMiller92JohnMiller92 Member
    edited September 2019

    @bikegremlin said:
    Click on a link at a website - wait one second for the new screen/page to load and be shown. That is what I meant. With locally run apps (not web/remote), it is easier to control/achieve.

    Hmm. I always thought less delay is better? I see what you mean though, but not sure.

  • @JustPfff
    Google's money are bigger, faster and stronger than Hetzner's money.
    In comparison, your money and mine as well, are only measurable in sacks of potatoes.
    And for them, we are mere communists, who congregate in front of the Low End store, drink beer and talk about life, stuff and other comradery.
    And that's why your server is slower than "big data".

    Thanked by 2uptime Ole_Juul
  • OP has not yet given any hint about the most basic issue anyone has asked about, which is ping time to Hetzner from where he or she is.

    Thanked by 3uptime Amitz ITLabs
  • ITLabsITLabs Member
    edited September 2019

    ^^^ 7

  • deankdeank Member, Troll

    OP is from a planet called Earth. That is all we need to know.

    Give it 500 years and we will be talking about pings between Earth and Mars.

    Thanked by 1JustPfff
  • OP please deliver!

  • bikegremlin said: it gives visitor an impression that something is happening.

    Yeah, like I'm standing at the front counter waiting for someone rummaging around in the back to find something they think I might like to see. Seriously, a 3 second delay for a site is unacceptable and shows lack of respect for viewers.

  • JustPfffJustPfff Member
    edited September 2019

    willie said: Did you compare your backend to a simple static page from the same server? What was the result?

    Here is the result txt file VS Nginx-Cache page using ApacheBench from my home connection :-
    ab -n 1000 -c 100 http://some.com/simple.txt || ab -n 1000 -c 100 http://some.com/somepage.html

    Document Path:          /simple.txt
    Document Length:        16024 bytes
    
    Concurrency Level:      100
    Time taken for tests:   42.404 seconds
    Complete requests:      1000
    Failed requests:        0
    Total transferred:      16293000 bytes
    HTML transferred:       16024000 bytes
    Requests per second:    23.58 [#/sec] (mean)
    Time per request:       4240.355 [ms] (mean)
    Time per request:       42.404 [ms] (mean, across all concurrent requests)
    Transfer rate:          375.23 [Kbytes/sec] received
    
    Connection Times (ms)
                  min  mean[+/-sd] median   max
    Connect:     1654 2552 390.4   2596    3351
    Processing:   844 1511 331.6   1501    6279
    Waiting:      844 1511 331.6   1501    6278
    Total:       2722 4063 468.4   4170    8455
    
    Percentage of the requests served within a certain time (ms)
      50%   4170
      66%   4320
      75%   4417
      80%   4463
      90%   4560
      95%   4652
      98%   4713
      99%   4752
     100%   8455 (longest request)
    ===================================================================
    
    Document Path:          /somepage.html
    Document Length:        44887 bytes
    
    Concurrency Level:      100
    Time taken for tests:   71.351 seconds
    Complete requests:      1000
    Failed requests:        0
    Total transferred:      45231152 bytes
    HTML transferred:       44887000 bytes
    Requests per second:    14.02 [#/sec] (mean)
    Time per request:       7135.052 [ms] (mean)
    Time per request:       71.351 [ms] (mean, across all concurrent requests)
    Transfer rate:          619.07 [Kbytes/sec] received
    
    Connection Times (ms)
                  min  mean[+/-sd] median   max
    Connect:     1872 2685 612.6   2585    8245
    Processing:  1889 4034 1401.2   3602   11352
    Waiting:      879 2245 1018.4   1904    7894
    Total:       4708 6719 1534.1   6187   15210
    
    Percentage of the requests served within a certain time (ms)
      50%   6187
      66%   6599
      75%   7025
      80%   7609
      90%   8803
      95%   9960
      98%  11507
      99%  12787
     100%  15210 (longest request)
    

    willie said: OP has not yet given any hint about the most basic issue anyone has asked about, which is ping time to Hetzner from where he or she is

    My current current connection is not the main issue , neither the ping delay , I already state it in OP page loading are not the same all the time their is dips between loading the pages (some loads fast and some load slow even I already visit it few seconds ago ) ,

    here is ping result from Hetzner ( BTW now I'm using 3G connection , but all time past days I was using 4G connection )

    ping 0.0.0.0
    PING 0.0.0.0 (0.0.0.0) 56(84) bytes of data.
    64 bytes from 0.0.0.0: icmp_seq=1 ttl=52 time=122 ms
    64 bytes from 0.0.0.0: icmp_seq=2 ttl=52 time=86.4 ms
    64 bytes from 0.0.0.0: icmp_seq=3 ttl=52 time=120 ms
    64 bytes from 0.0.0.0: icmp_seq=4 ttl=52 time=110 ms
    64 bytes from 0.0.0.0: icmp_seq=5 ttl=52 time=83.1 ms
    64 bytes from 0.0.0.0: icmp_seq=6 ttl=52 time=92.6 ms
    64 bytes from 0.0.0.0: icmp_seq=7 ttl=52 time=111 ms
    64 bytes from 0.0.0.0: icmp_seq=8 ttl=52 time=119 ms
    64 bytes from 0.0.0.0: icmp_seq=9 ttl=52 time=144 ms
    64 bytes from 0.0.0.0: icmp_seq=10 ttl=52 time=116 m
    
    Thanked by 1uptime
  • uptimeuptime Member
    edited September 2019

    @JustPfff said:

    Transfer rate: 375.23 [Kbytes/sec] received 
    

    [....]

    Transfer rate: 619.07 [Kbytes/sec] received 
    

    alrighty then ... thanks for that detail.

    Anyhoo - that's a fine kettle of fish - would it be possible to do that same analysis for the blogger site you are comparing to?

    A simple ping might also be helpful to compare. Maybe also traceroute or mtr.

    But it does seem like network to Hetzner may be an issue. I don't know that there is very much Hetzner can do about that - though possibly changing some routes or peering might help - but if you're some significant distance from them, it's a long shot.

    The thing is, I'm guessing you chose to use Hetzner because they offer nice specs for little money - not because they have a great network to your specific location. Ammirite?

  • williewillie Member
    edited September 2019

    JustPfff said: Here is the result txt file

    Thanks, the transfer speed looks terrible, like what you'd expect from an international connection in the early part of a transfer (since it is a small file). What is the ping time? Where are you geographically, if you are willing to say? If you are outside Europe, you should not expect good speed from a server that is in Europe.

  • I've started using Cloudflare Workers in front of my cached pages. The Workers run the WordPress site from Cloudflare's 160+ datacenters (locally) so my cache hit rate went to 98% from normally 70% and the site is loading super fast as Workers run nearest to the site visitor location! @JustPfff you should try it.

    Thanked by 2uptime ITLabs
  • @Sofia_K said:
    I've started using Cloudflare Workers in front of my cached pages. The Workers run the WordPress site from Cloudflare's 160+ datacenters (locally) so my cache hit rate went to 98% from normally 70% and the site is loading super fast as Workers run nearest to the site visitor location! @JustPfff you should try it.

    Any tutorials you might want to recommend? Is wordpress running on some backend and Workers are fetching it, or is the full stack running on workers? How much does it costs you (can it be covered by the free tier)?

  • DNS issue? if you go direct via your IP it should be instant load

  • @JohnMiller92 said:

    @bikegremlin said:
    Click on a link at a website - wait one second for the new screen/page to load and be shown. That is what I meant. With locally run apps (not web/remote), it is easier to control/achieve.

    Hmm. I always thought less delay is better? I see what you mean though, but not sure.

    To be fair, my info is about 2 decades old - things (might have) change(d). When I look at kids with smartphones swiftly scrolling - those brains might be a bit differently wired.

  • Sofia_K said: I've started using Cloudflare Workers in front of my cached pages. The Workers run the WordPress site from Cloudflare's

    Are these pages will be 100% cached or you could execute some dynamic codes . e.g I'm using rating plugin (even it's loaded from admin-ajax.php ) but I can execute wp-admin folder from page ruls , is that possiple with Cloudflare Workers .
    Plus my theme uses PHP condetion to detect mobile devices and load diffrent style for it , could I still do that with Cloudflare Workers ! I dont's want to do mobile redirect since it will make site indexing mass (I saw multiple sites Google were index their mobile subdomain instead of the main daomain ) .
    BTW I cold find soltion to disply diirent style form mobile using JavaScript but that will make pages more heavy for mobile users (since I'm hiding alot of HTML codes and big pictures )

  • SpeedBusSpeedBus Member, Host Rep

    Try putting the website behind CloudFlare? That would help cache a bit more of your website as well.

Sign In or Register to comment.