Recently our SugarCRM 7.6 server experienced an abnormally high CPU load. Additionally the outbound traffic increased tremendously. Here is what I've learned from logs.
- The daily bandwidth increased about 25 times.
- To 50% of the requests the server responded with 412 - Precondition Failed: Server does not meet preconditions
- The daily bandwidth increased due to GET requests made to /rest/v10/metadata?type_filter=currencies%2Cfull_module_list%2Cmodules_info%2Chidden_subpanels%2Cjssource%2Cjssource_public%2Cordered_labels%2Cmodule_tab_map%2Cmodules%2Crelationships%2Cserver_info%2Cconfig%2C_override_values%2Cfilters%2Clogo_url%2Ceditable_dropdown_filters&module_filter=&platform=base
- Requests to /rest/v10/metadata... constitute about 95% of the daily bandwidth.
Users reported that the application was slow, unresponsive or was loading forever.
I finally resolved the issue by doing a quick repair and rebuild. After that everything was back to normal.
Has anyone experienced anything similar? Why does SugarCRM suddenly start returning an HTTP status code 412 and why did a quick repair and rebuild help?