• If you are still using CentOS 7.9, it's time to convert to Alma 8 with the free centos2alma tool by Plesk or Plesk Migrator. Please let us know your experiences or concerns in this thread:
    CentOS2Alma discussion

Issue TTFB slow websites under plesk

garcue

Regular Pleskian
Hi everyone,

After updating plesk to version 18.0.30 I have started having problems with PHP. Currently I have loading times of my pages of more than 10 seconds when this did not happen before.

user@srv:~# curl -s -w '\nLookup time:\t\t%{time_namelookup}\nConnect time:\t\t%{time_connect}\nSSL handshake time:\t%{time_appconnect}\nPre-Transfer time:\t%{time_pretransfer}\nRedirect time:\t\t%{time_redirect}\nStart transfer time:\t%{time_starttransfer}\n\nTotal time:\t\t%{time_total}\n' -o /dev/null https://domain.com/wp-content/uploads/2016/11/img.png

Lookup time: 10,517745
Connect time: 10,522763
SSL handshake time: 10,542010
Pre-Transfer time: 10,542098
Redirect time: 0,000000
Start transfer time: 10,551202

Total time: 10,551225

Any ideas what I can do to correct the problem?
 
The numbers show that the only thing that is slowing down the site is the DNS lookup. Did you change anything on that, e.g. from IPv4 to IPv6? It almost "feels" like a timeout after 10 seconds. The system tries to resolve the address, e.g. to an IPv6 address, then it tries again to resolve to another address etc. What has been changed with DNS, IP addresses or things that are related to it?
 
Hi Peter,

Thanks for your answer, we haven't touched anything regarding DNS or IP. I have other servers configured in the same way and those work without problem.

I've started having problems after updating to plesk obsidian 18.0.31.

I don't know if someone else has the same problem.

All the websites that I have under the server load with a TTFB of at least 10 seconds.

When a week ago they worked perfect
 
Last edited:
I just ran a test on pingdom and this has been the result, so we downloaded the DNS issue

1605649726389.png
 
That's interesting, because it does not at all match the first measurement.

What DNS servers is your system using (listed in /etc/resolv.conf)? Have you tried using other servers there, just to rule out that the system might have a DNS resolution issue? Have you checked what IP protocol version is currently being used in the domain, e.g. if you have set IPv6 and IPv4 but the server only uses IPv4?

I am asking for DNS again, because I know the issue basically from FTP where some FTP software tries to connect through IPv6 first if available, then fails, and several seconds later does a new attempt with IPv4, which leads to exactly this constant delay that you are experiencing with your website. It might be a similar issue, because in your initial post the lookup time was almost equal the transfer time.
 
Back
Top