Magento2 Optimised Stack Troubleshooting Guide¶

Magento2 Log Location:¶

Magento2 will capture errors and store them in files within the report directly typically located:

/var/www/vhosts/example.com/htdocs/var/report

If you’re having an issue and there is no files generated in the report directory then review the exception.log and system.log files within var/log.

/var/www/vhosts/example.com/htdocs/var/log

PHP-FPM error log location:¶

Default error logs: /var/log/php-fpm/error.log
Site error logs: /var/www/vhosts/example.com/logs/example.com-phpfpm-error.log

Multi instance default error logs: /var/opt/remi/$MUTLIPHPVERSION/log/php-fpm/error.log
Multi instance site error logs: /var/www/vhosts/example.com/logs/example.com-phpfpm-error$PHPVERSION.log

Nginx log location¶

Site access logs:

/var/www/vhosts/example.com/logs/$DOMAIN-access.log

Or

/var/log/nginx/example.com-access.log

Site error logs:

/var/www/vhosts/example.com/logs/example.com-error.log

Or

/var/log/nginx/example.com-error.log

MySQL error log¶

You can find the location of the mysql logs using the following command:

~]# mysql -e "SHOW VARIABLES LIKE 'log_error'"
+---------------+---------------------+
| Variable_name | Value               |
+---------------+---------------------+
| log_error     | /var/log/mysqld.log |
+---------------+---------------------+

Elasticsearch error log¶

The typical Elasticsearch log location is here:

/var/log/elasticsearch/elasticsearch.log

Server Logs¶

You can check for server errors here:

/var/log/messages
/var/log/crash
/var/log/dmesg
/var/log/syslog

Varnish Logs¶

Varnish logs are not written by default. In order to get the log data you will need to run the following command:

varnishlog

You can find more information on varnishlog here: https://docs.ukfast.co.uk/ecommercestacks/magento/magento2/varnish/varnish.html#varnishlog

Command to check Nginx access logs:¶

grep "`date +%d/%b/%Y`" /var/log/nginx/example.com-access.log | awk '{print $1, $5,$6, $7,$9 $11}' | awk -F\" '{print $1,$2,$3}' | awk '{print $1,$3,$4,$5}' | sort | uniq -c | sort -gr | head -n 20

The most common cause of server slow down is one or more processes consuming a large amount of the system resources. The easiest way to view the current usage of resources is to use the utility top. top allows you to view CPU utilisation, Memory usage, Disk activity, Tasks and Load Average. Many of these are quite self explanatory and percentages to show usage. Load Average is representative of the queue of instructions on the CPU, this should ideally remain below 1.

There are several variations on the top utility however those are not covered in this section.

PHP-FPM Max Children Reached¶

Run the following command to check if any php-fpm instance has reached max_children on CentOS or Ubuntu:

for m in $(for i in $(ps awux | grep -w "[p]hp-fpm.conf" | awk '{print $NF}' | sed 's/^.//' | sed 's/.$//'); do grep error_log $i; done | awk '{print $NF}'); do grep -H "reached pm.max_children" $m; done

If max_children has been reached reload php-fpm and then review the server resources using atop to define if the max_children value should be changed.

Database Deadlocks¶

You can get the engine status of MySQL using the following command:

mysql -e "SHOW ENGINE INNODB STATUS;"

This will identify if there has been a deadlock

Varnish 503¶

You can check the health of the application using the following command:

~]# varnishadm backend.list
200
Backend name   Admin    Probe    Health
boot.default   probe    10/10    healthy

This is the typical configuration for the healthcheck: https://docs.ukfast.co.uk/ecommercestacks/magento/magento2/varnish/varnish.html#health-check

Permissions¶

Make sure the owner and group of the document root is "websiteuser:websiteuser". You can find the user and group from the PHP-FPM configuration pool file.

find /var/www/vhosts/sitename.co.uk/htdocs/ -! -user websiteuser

chown -R websiteuser:websiteuser /var/www/vhosts/sitename.co.uk/htdocs/