Asterisk version 18.26.4, FreePBX Version 15.0.37.9
When accessing the server via its web interface, it is extremely slow to load modules and tabs. Often times, I have to click the button 3-4 times to elicit any sort of response. When working on a template in Endpoint Manager, attempting to save changes crashes the webpage or it becomes unresponsive.
This issue first appeared 2-3 weeks ago. Nothing had changed on the server prior to that aside from routine module and system maintenance. It is running on a Sangoma FreePBX 400 “Blue Box” server.
Any tips or tricks on resolving this? We do have eventual plans to migrate this instance over to a Virtual Machine around Christmas Break.
Affirmative. Nothing out of the ordinary is listed in any of the checks. Our system appears clean.
We first started noticing this issue on August 19th.
Update. Restarted the system and ran module updates again. No change. In fact, it seems to have gotten even worse!
This is a system with about 160 extensions and all of the resources seem fine. I double-checked the vulnerability checklist and our system is still clean (it’s only accessible from a few IPs within our network). I’m half tempted to rebuild it from scratch, but that’s not feasible for at least a month.
Not sure if it would be the same issue, but when I built out a system for a school (close to 900 extensions), it would get slower the more phones came online. It was an issue with not enough threads on mpm_prefork and it kept crashing.
Check the logs here: /var/log/httpd/error_log
Look for something like the following: [mpm_prefork:error] [pid 1786] AH00161: server reached MaxRequestWorkers setting, consider raising the MaxRequestWorkers setting
If that’s showing up, then it’s capping out. Currently, my system in question has 12 cores assigned and 24gb ram. On CentOS, I found the config here: /etc/httpd/conf.modules.d/00-mpm.conf
Debian 12 with FreePBX17 I found it here: /etc/apache2/mods-available/mpm_prefork.conf
This may be totally unrelated to what you find, but changing that did have a big impact for me. Restart your system after setting this. Just restarting httpd or apache2 didnt seem to have full affect.
If it’s just the UI and asterisk is still running fine then you may need to adjust your apache settings as per the previous post. Also make sure you are not running out of memory and/or using a lot of swap memory.
It’s the UI, but mostly Endpoint Manager and the UCP at this point. Other modules do freeze up still, but Endpoint Manager is pretty much useless at this point. The UCP refuses to load at all. Module Admin, especially when I attempt to update modules, is also an issue. Endpoint version is 15.0.66.
We did check and disable a few ports that were inadvertently left open. After running a few more sanity checks, our machine still appears clean and is still running slow.
I was able to find a solution. Either a certificate in FreePBX changed or something within Google Chrome changed, and the TLS handshake was not working. I changed to using SeaMonkey as my browser and there are no longer any issues.