We launched new forums in March 2019—join us there. In a hurry for help with your website? Get Help Now!
    • 34127
    • 135 Posts
    Quote from: rthrash at Feb 21, 2007, 11:58 AM

    How did you think we were planning on implementing the coffee bit? tongue
    No more having to get up to make my own coffee anymore! Perfect for those late-night coding sessions. laugh
      • 28042 ☆ A M B ☆
      • 24,524 Posts
      Ah, the light flashes on at last! Now I know where the "sticky bit" comes from! Somebody implemented a "coffee bit", and of course spilled it on his keyboard, and voila... sticky bit! grin

      (for the UNIX-impaired: http://en.wikipedia.org/wiki/Sticky_bit)
        Studying MODX in the desert - http://sottwell.com
        Tips and Tricks from the MODX Forums and Slack Channels - http://modxcookbook.com
        Join the Slack Community - http://modx.org
        • 28331
        • 10 Posts
        Hello, and thanks for the answers. Hope to see a newer version soon then smiley

        But I’m still wondering if it’s possible to do a workaround as i wrote above:

        // ugly pseudocode
        <? import ’/plugins/myplugin.php’ ?>

        instead of pasting all code inside the MODx, and then forcing the MODx-engine to NOT store any cached phpcode inside of siteCache.idx.php.

        I know it can be done to import php-scripts, but I’m not sure how the engine works. Would it still cache php and make sitecache.idx.php grow, or just leave it? This way I could temporary (since 0.9.6 is not out yet Smiley ) make the sitecache-file much smaller, and then get better result for my benchmarks.
          • 15987
          • 786 Posts
          I do that all the time with my snippets, especially when doing development on them. I’m not sure about plugins but I would think it would work there as well.
            • 20400
            • 37 Posts
            Just thought I would add my two cents.

            I ran http_load against Modx on my test server and got similar results:

            http_load -parallel 5 -fetches 100 /home/mregan/bin/test.url
            100 fetches, 5 max parallel, 773600 bytes, in 5.45256 seconds
            7736 mean bytes/connection
            18.34 fetches/sec, 141878 bytes/sec
            msecs/connect: 0.15374 mean, 0.237 max, 0.027 min
            msecs/first-response: 271.273 mean, 5452.32 max, 53.746 min
            HTTP response codes:
            code 200 -- 100

            However, I thought that was pretty good. When I run it against the same code on a shared host machine I get much slower response.

            You can’t blame the Modx cache for all the delays. (Although it is a bit of a bear.) There are two other bottle necks the .htaccess and the start time for PHP.

            Anyway - just a thought.

            Some one asked about other CMS respose times. I ran the same test against WordPess on the same test server - just the intro I don’t have any entries - it was 10 fetches/sec
              • 34127
              • 135 Posts
              I actually decided to run the benchmark on my own server to see if I got similar results. I’m running an Opteron 246, 2GB RAM, Apache 1.3 and PHP 5.1. Fresh MODx install too:

              Concurrency Level:      10
              Time taken for tests:   23.062 seconds
              Complete requests:      1000
              Failed requests:        0
              Broken pipe errors:     0
              Keep-Alive requests:    0
              Total transferred:      13845000 bytes
              HTML transferred:       13534000 bytes
              Requests per second:    43.36 [#/sec] (mean)
              Time per request:       230.62 [ms] (mean)
              Time per request:       23.06 [ms] (mean, across all concurrent requests)
              Transfer rate:          600.34 [Kbytes/sec] received
              
              Connnection Times (ms)
                            min  mean[+/-sd] median   max
              Connect:        0     0    0.6      0    16
              Processing:    41   230  194.4    187  2200
              Waiting:       40   229  194.4    187  2200
              Total:         41   230  194.4    187  2200
              
              Percentage of the requests served within a certain time (ms)
                50%    187
                66%    225
                75%    256
                80%    283
                90%    372
                95%    558
                98%    811
                99%    985
               100%   2200 (last request)


              Edit: and just for fun, I threw a quick install of WP on my server and benchmarked it for comparison:

              Concurrency Level:      10
              Time taken for tests:   60.534 seconds
              Complete requests:      1000
              Failed requests:        0
              Broken pipe errors:     0
              Keep-Alive requests:    0
              Total transferred:      4925262 bytes
              HTML transferred:       4714051 bytes
              Requests per second:    16.52 [#/sec] (mean)
              Time per request:       605.34 [ms] (mean)
              Time per request:       60.53 [ms] (mean, across all concurrent requests)
              Transfer rate:          81.36 [Kbytes/sec] received
              
              Connnection Times (ms)
                            min  mean[+/-sd] median   max
              Connect:        0     0    1.6      0    30
              Processing:   169   600  503.4    544  9887
              Waiting:      168   599  503.4    544  9886
              Total:        169   600  503.4    544  9887
              
              Percentage of the requests served within a certain time (ms)
                50%    544
                66%    617
                75%    664
                80%    701
                90%    814
                95%    932
                98%   1169
                99%   1449
               100%   9887 (last request)
              


              By that, MODx is nearly 3 times as fast as Wordpress? shocked
                • 26085
                • 494 Posts
                I try to test modx 096 with HttpTrafficGen.
                And reseived such result


                1 user, without HttpTrafficGen:
                MySQL: 0.0066 s, 0 request(s), PHP: 0.2340 s, total: 0.2407 s, document retrieved from cache.
                MySQL: 0.0702 s, 17 request(s), PHP: 0.5033 s, total: 0.5735 s, document retrieved from database.

                11 users at the same time:(Me + 10 HttpTrafficGen)
                MySQL: 1.7143 s, 17 request(s), PHP: 2.2653 s, total: 3.9796 s, document retrieved from database.
                MySQL: 0.3066 s, 0 request(s), PHP: 1.1193 s, total: 1.4259 s, document retrieved from cache.

                I think it’s quite good.

                p.s. my test pc P3 700, 196mb,
                WinXP SP2, apache 1,3 php 4, mysql 5
                  • 28331
                  • 10 Posts
                  Quote from: Stabilovic at Feb 20, 2007, 10:35 AM

                  Hello!
                  I’ve been using ModX a couple of months now and really like it a lot.
                  My main consern though, is that the cms seems to be very slow.

                  Okay!
                  I apologize for that statement above, since you’ve showed me the Wordpress benchmarks smiley

                  If I rephrase myself then. Do you think that ModX can handle a couple of thousand visitors at the same time? No logins. Just clicking, reading and searching.
                  We’re going to integrate some forum in the design too, but I’m not sure if it’s going to be SMF or vBulletin for now.

                  I believe the server is going to be a very fast Opteron cpu, 2 gigs of ram and scassi-drives.


                    • 34127
                    • 135 Posts
                    I’m not sure if I’m qualified to answer this, and a MODx developer could probably give a more definitive answer, but it depends on several things like how many snippet calls you have and how database intensive they are, and how active you visitors are. If you have a bunch of snippet calls on every page, all accessing the database to pull records of some sort, and if everyone is searching (which is generally very database intensive in itself), any app would have a problem. The bottleneck is transfer: with enough RAM and a fast enough harddrive, I’m sure it could work. IF you don’t have tons of crazy snippets all accessing the database on every page load, and if you have enough RAM and your harddrive is fast enough (perhaps a RAID5 setup), I don’t see why MODx couldn’t handle it. It’s the same deal with forums like vB... the software can handle it, but the hardware is the bottleneck. smiley


                    On the topic of benchmarks, I think it would be sweet if we could get some speed comparisons up to show how fast MODx really is compared to other CMSes. I mean, WP is supposed to be fast and lightweight (it’s a blogging tool, right?) and MODx DESTROYS it. I’m sure MODx wouldn’t have much of a problem tearing Joomla a new one. tongue
                      • 28373
                      • 204 Posts
                      Quote from: Stabilovic at Feb 26, 2007, 03:13 AM

                      Quote from: Stabilovic at Feb 20, 2007, 10:35 AM

                      Hello!
                      I’ve been using ModX a couple of months now and really like it a lot.
                      My main consern though, is that the cms seems to be very slow.

                      Okay!
                      I apologize for that statement above, since you’ve showed me the Wordpress benchmarks smiley

                      If I rephrase myself then. Do you think that ModX can handle a couple of thousand visitors at the same time? No logins. Just clicking, reading and searching.
                      We’re going to integrate some forum in the design too, but I’m not sure if it’s going to be SMF or vBulletin for now.

                      I believe the server is going to be a very fast Opteron cpu, 2 gigs of ram and scassi-drives.




                      We have modx powering a high traffic adult site - visitors to the site, view a substantial volume of high res images, access hundreds of content pages, and use several dozen types of contact forms. The site takes 90,000 hits a day, it’s hosted on a shared server plan over at hostgator. It eats on average, 19gb in bandwidth each month.

                      The manager isn’t as fast as we’d like, but the site has been ranked as ’faster than average’. We did hit problems with the stats module, we turned that off, and moved to analytic’s.

                      The site has been in operation for 9 or so months, no crashes, no complaints.