DHCP lease - 504 Gateway Time-out

empbilly

Hello,

When accessing the dhcp lease, error 504 Gateway Time-out occurs. I've tried some of the options listed in the topics I found with the same problem, but as they are older topics, I believe they have already been fixed.

I'm using version 2.7.2
Two appliances with HA CARP
The DNS configured in System > General Setup is our local authoritative DNS
DNS forwarder active on some interfaces

The error in Status > System Logs > System > General is:

2024/05/09 20:41:43 [error] 86295#100359: *110120 upstream timed out (60: Operation timed out) while reading response header from upstream, client: xxx.xxx.xxx.xxx, server: , request: “GET /status_dhcp_leases.php HTTP/2.0”, upstream: “fastcgi://unix:/var/run/php-fpm.socket”, host: “pfmaster.sertao.ifrs.edu.br”, referrer: “https://pfmaster.sertao.ifrs.edu.br/”

lease file

ls -laht dhcpd.leases
-rw-r--r-- 1 dhcpd _dhcp 2.1M May 9 20:44 dhcpd.leases

grep -o -i 'lease' dhcpd.leases | wc -l
    18967

Anything else I can provide to help solve the problem?

EDIT:
###### After resetting dhcpd.leases on both appliances following the wiki and rebooting, the dhcp lease page is ok again.

johnpoz

@empbilly your leases file was 2.1M?

I was just curious what mine was ;)

[23.09.1-RELEASE][admin@sg4860.home.arpa]/var/dhcpd/var/db: ls -laht dhcpd.leases
-rw-r--r--  1 dhcpd _dhcp   18K May  9 20:10 dhcpd.leases
[23.09.1-RELEASE][admin@sg4860.home.arpa]/var/dhcpd/var/db: grep -o -i 'lease' dhcpd.leases | wc -l
      74
[23.09.1-RELEASE][admin@sg4860.home.arpa]/var/dhcpd/var/db:

What is it now?

empbilly

@johnpoz said in DHCP lease - 504 Gateway Time-out:

your leases file was 2.1M?

At first, yes.

Does the lease file only contain the IPs that have been granted or does it list all the possible IPs configured in the dhcp of each vlan?

johnpoz

@empbilly it should only have actual leases that have been handed out.

Take a look for example - here is lease my applewatch just got

lease 192.168.2.221 {
  starts 5 2024/05/10 11:57:58;
  ends 6 2024/05/11 11:57:58;
  cltt 5 2024/05/10 11:57:58;
  binding state active;
  next binding state free;
  rewind binding state free;
  hardware ethernet 7c:61:30:ad:55:47;
  uid "\001|a0\255UG";
  client-hostname "JohnsAppleWatch";
}

Guess you could of something going bonkers and just kept grabbing new lease all the time.. But 18k some leases, is that the size of your network - do you have anything even close to that many devices on your network.. Or was something just grabbing up leases like crazy?