Hello, This morning at 10:00 I updated tha information of a user in ispconfig to see how it updated the /etc/passwd file. It's now 18:30 and the process still running This is a ps -ef report of the affected processess: ****************************************** root 3850 3734 41 13:04 ? 02:17:45 /root/ispconfig/php/php -q /root/ispconfig/scripts/writeconf.php mysql 3196 3163 3 01:02 ? 00:38:15 /usr/libexec/mysqld --defaults-file=/etc/my.cnf --basedir=/usr --datadir=/var/lib/mysql --user=mysql --pid-file=/var/run/mysqld/mysq ld.pid --skip-locking --socket=/var/lib/mysql/mysql.sock ****************************************** This machine is a HP DL380G4 with DUAL Xeon, 3 Gb of RAM and a SAN (MSA1000) by fiber for disk. I have checked, and optimied the database. The process seems to be in a "loop". What can I do? Thank's in advance.....Larry
What's the output of Code: ls -la /root/ispconfig ? Any warnings in /home/admispconfig/ispconfig/ispconfig.log?
Hello again, Finally I find out what was going on. First of all I want to point out that the server has to process the following data each time I change a user: 1 Client. 3 Sites 1 of the sites with 3000 users and 28 co-domains. What happens is that the script ispconfig_sendmail.lib.php took way too long to process the co-domains, consuming lots of "mysql" resorces. I find out by inserting entry-logs in the code. Finally a solved the problems using 2 aproaches: 1) I changed the select from the co-domains to be done only once per site. ********************************************* if (empty($codomains)){ $codomains = $mod->db->queryAllRecords("SELECT * from isp_dep,isp_isp_domain where isp_dep.child_doc_id = isp_isp_domain.doc_id and isp_dep.child_doctype_id ='".$isp_web->domain_doctype_id."' and isp_dep.parent_doctype_id = '".$isp_web->web_doctype_id."' and isp_dep.parent_doc_id = '".$web_doc_id."' and isp_isp_domain.status != 'd'"); } if(!empty($codomains)){ foreach($codomains as $codomain){ ********************************************* 2) I changed the search to be done only on those domains with the dns-mx checked. ********************************************* if (empty($codomains)){ $codomains = $mod->db->queryAllRecords("SELECT * from isp_dep,isp_isp_domain where isp_dep.child_doc_id = isp_isp_domain.doc_id and isp_dep.child_doctype_id ='".$isp_web->domain_doctype_id."' and isp_dep.parent_doctype_id = '".$isp_web->web_doctype_id."' and isp_dep.parent_doc_id = '".$web_doc_id."' and isp_isp_domain.status != 'd' and isp_isp_domain.domain_dnsmail = '1'"); } if(!empty($codomains)){ foreach($codomains as $codomain){ ********************************************* With these changes, I reduced the process time dounw to 1 hour, and everything finished ok. I would like to know your point of view, mainly to see if this was the best aproach, or if you think of a better solution to change this one. Normally a domain can have lot's of co-domains, so all of them point to the same web. That fact not always works for mail addresses, so the size and the time expent to create the "/etc/mail/virtusertable" can be reduced a lot. Thank's in advance.....Larry.
