1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Debug reporting higher memory usage on a different server, same files/db

Discussion in 'Server Configuration and Hosting' started by rainmotorsports, Apr 28, 2014.

  1. rainmotorsports

    rainmotorsports Well-Known Member

    Been having a long time issue with my current host ever since they upgraded to PHP 5.4 that anytime an action that sends an email causes the resulting page to error, timeout, double request or connection reset. (No php errors thrown, no access to the apache logs...).

    Our current server debug mode reports 6 to 7 mb memory usage. I threw up an Apache server with Percona 5.6 tried both PHP 5.3 and 5.4. Its reporting 19 to 20mb a page. Is this being caused by the lack of some installed modules using less efficient fallbacks or is the memory used more related to server configuration and not Xenforo? I guess the basic issue is that I don't know what that memory usage measurement is of.
  2. Floren

    Floren Well-Known Member

    Sorry but I did not understood nothing. All I got is that you have errors when an email action is performed. Yet you say that you get no PHP and Apache errors... does that makes sense? We don't even know what OS you run... Did you looked at mail queue to see if is sending out your mail? Do a basic test:
    At least you will know if your mail server works properly. For the rest, I doubt someone will reply as nothing is clear, beside "throwing" a server. If you did that, then the default client size is 1MB, not 20.
  3. Floren

    Floren Well-Known Member

    All you tell us is: the site is reporting 19 to 22 MB of memory on a page load versus 6 or 7
    Do yo really think is clear what you are saying or enough to help you figure what is wrong with your setup?

    I'll need your help also to figure why my bike's fuel injection acts weird. Should be enough for you to figure it and let me know exactly how to fix it. :giggle:
  4. Rigel Kentaurus

    Rigel Kentaurus Well-Known Member

    Did by any chance you jump from a 32bit architecture to 64bit?
  5. rainmotorsports

    rainmotorsports Well-Known Member

    Crap I didn't see your second reply when I went to delete that post I was rewriting it and was going to post again.

    As for your example, you already have isolated the issue further than I have. I don't even know what the memory usage is for is it php? Your bikes not performing well and you have already determined its not intake or ignition where down to the fuel system which has 3 major functioning components. Fuel Supply, Pressure Regulation, The Injectors and their controller. All of which have pretty obvious individual tests.

    Well I told you the best I could about the server setup, I can add a little more actually. As for the server I am at square one. I haven't configured anything its on the absolute default settings for the Percona config for example, I dont have a my.cnf yet I have done no tuning. The install is on a tuned host and then I move it to a barebones apache 2.2, php 5.3 install with Percona 5.6. If I went in and generated non standard configs and installed extra modules like ImageMagick I could tell you that but Ive done... rather almost nothing.

    Current Host - 64 bit Gentoo shared host running Mysql 5.5 php 5.4,
    Test Host 64 Bit Ubuntu 12.04 LTS, webmin installed apache and php 5.3, stripped the script for mysql and installed percona myself. Modified the startup script to work with webmin. Currently no tuning, non required configs. I am not using a cache for xenforo.

    I am renting a VPS from WSDW today see what the memory usage is on it once they get the control panel installed. Basically I am not "worried" about the memory usage so much as to the general cause if I go on one host and then another it would seem triple the usage would be something to worry about when tuning later though.

    I know for a fact that the test VPS is struggling for as lack of memory at the moment which is causing the load times to be higher than normal. I no longer intend to use this particular server.

    I just checked and uname -m is reporting 64 bit kernel on the current host, the test server is 64 bit as well. That would have made some sense though.
    Last edited: Apr 29, 2014
  6. Rigel Kentaurus

    Rigel Kentaurus Well-Known Member

    In a page with phpinfo() check the "Configure Command" section (./configure --prefix=/usr ... etc)
    There are different flags at compile time that may enable modules.

    Also, check your php.ini for any extensions that are enabled

    PHP extensions have a lifecycle, and most of them get loaded into memory or load data into memory. That means that if one box has a lot of extensions (imap, sockets, php, gd, imagick, etc), then a single request would be a lot heavier.
  7. rainmotorsports

    rainmotorsports Well-Known Member

    Oh this is going to be fun. So the memory usage reported at the bottom of the board is being reported by PHP for that page load?

    Ran phpinfo on both servers the webmin based test server is reporting nothing for configure command, but my current host has plenty in there:
    './configure' '--prefix=/usr' '--build=x86_64-pc-linux-gnu' '--host=x86_64-pc-linux-gnu' '--mandir=/usr/share/man' '--infodir=/usr/share/info' '--datadir=/usr/share' '--sysconfdir=/etc' '--localstatedir=/var/lib' '--prefix=/usr/lib64/php5.4' '--mandir=/usr/lib64/php5.4/man' '--infodir=/usr/lib64/php5.4/info' '--libdir=/usr/lib64/php5.4/lib' '--with-libdir=lib64' '--without-pear' '--disable-maintainer-zts' '--enable-bcmath' '--with-bz2=/usr' '--enable-calendar' '--enable-ctype' '--with-curl=/usr' '--without-curlwrappers' '--enable-dom' '--without-enchant' '--enable-exif' '--enable-fileinfo' '--enable-filter' '--enable-ftp' '--with-gettext=/usr' '--with-gmp=/usr' '--enable-hash' '--without-mhash' '--with-iconv' '--disable-intl' '--enable-ipv6' '--enable-json' '--without-kerberos' '--enable-libxml' '--with-libxml-dir=/usr' '--enable-mbstring' '--with-mcrypt=/usr' '--with-mssql=/usr' '--with-onig=/usr' '--with-openssl=/usr' '--with-openssl-dir=/usr' '--disable-pcntl' '--enable-phar' '--enable-pdo' '--with-pgsql=/usr' '--enable-posix' '--with-pspell=/usr' '--without-recode' '--enable-simplexml' '--disable-shmop' '--without-snmp' '--enable-soap' '--enable-sockets' '--with-sqlite3=/usr' '--without-sybase-ct' '--disable-sysvmsg' '--disable-sysvsem' '--disable-sysvshm' '--without-tidy' '--enable-tokenizer' '--disable-wddx' '--enable-xml' '--enable-xmlreader' '--enable-xmlwriter' '--with-xmlrpc' '--with-xsl=/usr' '--enable-zip' '--with-zlib=/usr' '--disable-debug' '--enable-dba' '--without-cdb' '--with-db4=/usr' '--disable-flatfile' '--with-gdbm=/usr' '--disable-inifile' '--without-qdbm' '--with-freetype-dir=/usr' '--with-t1lib=/usr' '--disable-gd-jis-conv' '--with-jpeg-dir=/usr' '--with-png-dir=/usr' '--without-xpm-dir' '--with-gd' '--with-imap=/usr' '--with-imap-ssl=/usr' '--with-interbase=/usr' '--with-ldap=/usr' '--with-ldap-sasl=/usr' '--with-mysql=mysqlnd' '--with-mysql-sock=/var/run/mysqld/mysqld.sock' '--with-mysqli=mysqlnd' '--with-unixODBC=/usr' '--with-pdo-dblib' '--with-pdo-mysql=mysqlnd' '--with-pdo-pgsql' '--with-pdo-sqlite=/usr' '--with-pdo-odbc=unixODBC,/usr' '--with-readline=/usr' '--without-libedit' '--without-mm' '--with-pic' '--with-pcre-regex=/usr' '--with-pcre-dir=/usr' '--with-config-file-path=/etc/php/apache2-php5.4' '--with-config-file-scan-dir=/etc/php/apache2-php5.4/ext-active' '--disable-embed' '--disable-cli' '--disable-cgi' '--disable-fpm' '--with-apxs2=/usr/sbin/apxs'
    I really don't see any additionally loaded extensions in eithe php.ini. The CGI handler seems to be one of the biggest differences. The test server is running FastCGI while I am not really familiar with the current hosts setup its using is showing as "Apache 2.0 Handler".

    Im guessing its FastCGI using so much more memory? Guess I might have to get my hands dirtier than I wanted to. Webmin kinda sucks anyways, still makes things a little easier with all the crap I have going on.

    EDIT - Guess not, handler makes no difference on the reported memory usage though of course actual memory usage total server side is still going to be a different story.
    Last edited: Apr 29, 2014

Share This Page