0

We have a Debian box that has Apache installed, whose job is mainly (also serves some locally hosted PHP files) serving as a reverse proxy for Tomcat based web apps that run on other servers.

From time to time we either get an alert from UptimeRobot or a call from one of our customers (sometimes even both) telling us that the system is down. We haven't been able to pinpoint the root cause of this problem but, the only evidence that we've observed is that ssh access to the server doesn't work when in the "system down" scenario, so something else asides from the Apache server must get clogged/exhausted, perhaps from the network connections point of view.

Software versions:

>lsb_release -a
No LSB modules are available.
Distributor ID: Debian
Description:    Debian GNU/Linux 8.2 (jessie)
Release:    8.2
Codename:   jessie

>apachectl -v
Server version: Apache/2.4.10 (Debian)
Server built:   Aug 28 2015 16:28:08

What other info, config or perhaps log files should I provide to make diagnostic easier?

gvasquez
  • 153
  • 8
  • 1
    Set up some pro-active monitoring of services/resources as well as log files over time. This may not help you immediately but will give you more information regarding possible deterioration of resource over time. – Raman Sailopal Jul 03 '17 at 20:42
  • @RamanSailopal we already have some metrics being monitored (CPU, memory, and storage), and apache his own logs files. But which specifically would you suggest to add/enable? – gvasquez Jul 03 '17 at 21:08
  • 1
    I use Nagios and Graylog2 but there are many different options. I would also set up fail2ban as well to discount any possible ddos scenarios. – Raman Sailopal Jul 03 '17 at 21:19
  • @RamanSailopal fail2ban is enabled, but as request come through CloudFlare, I guess it's not really useful. Regarding both Nagios and Graylog they are good tools, but what metrics to gather with them? – gvasquez Jul 04 '17 at 01:48
  • Nagios-graph is a good option. It's a bit fiddly especially if you are not used to Perl but it does a good job. – Raman Sailopal Jul 04 '17 at 07:47
  • 1
    Did you check the other logs, such as mysql and php, or whatever else you are using? Did you check the resources? It's possible that it's crashing because of memory issues and not apache issues. See if your host has information on CPU use and crash history. – Christia Oct 02 '17 at 22:13

0 Answers0