Check environment resources states:
Examine the % of Storage Free and MySQL % free storage by node frames.
Examine the % of free system memory and Swap memory free in bytes frames.
Examine the Alerts during the timeframe frame.
Examine the CPU % by host frame: If it is exhibiting high CPU utilization, check the Support Knowledge Base article in the header for the frame. Also, check to ensure that database imports/exports or backups are not being done during peak traffic periods.
Check the Web Traffic volume compared to one week ago frame: If the traffic is much higher than the prior week during the same period, can it be explained (sale campaign or new products that have been marketed, for example)?
Site Down?
link in the frame header. The frame will identify any errors that are occurring and their frequency.The IP Frequency tab identifies request frequency from IP addresses that are made against the origin servers (meaning that the request could not be served from Fastly as 74 it wasn’t cached).
Check the Database Errors frame for errors that may coincide with the issue/problem timeline.
Check the Database mysql-slow.log frame to identify SQL statements that are occurring. INSERT
, UPDATE
, and DELETE
commands may take a while if the query is not optimized. Even SELECT
statements can be very inefficient if done against large tables.
PHP States and PHP Errors frames will show potential issues with PHP. The PHP States frame will show PHP process terminations, starts, and when the service reaches the ready state by node. The PHP Errors frame may help isolate where the issue is with PHP, such as memory size, workers, or the number of servers.
To see latency in transactions, the Transactions – Avg, Max, Min table can be sorted by column to show the longest-running transaction duration. An overloaded cluster will have latent durations in transactions, but it will also show anomalies that might pinpoint a problem with a method or cron.
The Cron error frame will show cron locks, SQL errors that may be associated with cron logs, and shared staging crons that may be running on production environments when there is a dedicated staging environment.
The ElasticSearch Errors frame shows errors that may indicate major problems with Elasticsearch queries, data, or indices.