Time to First Byte Issues on Shared

Time To First Byte, or TTFB, refers to the time the site takes to completely resolve and begin sending data.

This is usually indicated in both the Developer Console of browsers and GTmetrix as the first possible delay in a waterfall readout. This can result in the website loading slowly before any resources are loaded.

Confirm No Server-Side Usage Issues

1. SSH into the appropriate server
2. Run uptime and record load averages
   EXAMPLE:
   

   [userna5@ecngx100: ~]$ uptime
   14:38:23 up 777 days, 12:31, 6 users, load average: 22.73, 24.57, 25.89

3. Run nproc to find how many cores are present on the server
   EXAMPLE:
   

   [userna5@ecngx100: ~]$ nproc
   48

4. Using the nproc result, determine whether the server at large is under high load as compared to the nproc result
   1. If the load average result is equal to the nproc result, the server is at 100% usage
   2. Likewise, if the load average result is half the nproc result, the server is at 50% usage
      NOTE: Using the results of previous examples, we can determine that this server has been hovering well under 50% usage (20 uptime load average / 48 cores), and is, therefore not under heavy load.
      TIP: Our Systems team constantly monitors the load on all shared servers and takes actions to drop the load.  In general, the load on the server is considered elevated if it's higher than 40 for more than 20 minutes, higher than 60 for more than 5 minutes, or higher than 100 for one minute.
5. If no heavy load is present, continue on to the next section

No Server-Wide Load Issues

1. Run top c -u userna5 to get a current usage readout for the user
   REPLACE: userna5 with the cPanel username of the domain.
   TIP: Top processes can be taxing on the server itself, in general, this is nothing to worry about. If, however, the situation calls for that kind of concern, you can run ps aux | grep userna5 instead.
   NOTE: The C argument allows a full readout of what processes are running. If the results go off the terminal screen, the arrow keys can be used to scroll through the results.
2. If any results show heavy usage, it can provide an indication of what processes may be causing the issue
3. Otherwise, it's advised to work with a developer to optimize the site in question