new device shows "500 internal server error"

[GGA]

I would like to add a UPS/netman plus to our v1.2.22 Cacti setup.
But adding show me the following error message:
Sorry, we could not process your last action.
Error:
500 Internal Server Error
Reason:
SNMP Information
There must be something unexpected with this UPS. But I can't see what.
The same UPS is working on our old 0.8.8b Cacti.
All other UPS are working on the v1.2.22. Well they all use different network controller.
I can do a snmpgetnext -v2c -c<community> <ip> .1.3 and I receive an answer. Not the typically sysDescr.0, would that be important? But this behavior is not new, it allways worked liked this.

What could I do to go deeper? There is a device debug switch but there seems nothing specific written in any cacti-logfile.
Thanks for any advice.

[TheWitness]

What PHP version? PHP has been making lots of fun for us in recent years. What page? Any plugins?

[GGA]

Ah sorry for the lack of informations.
Its a generic Ubuntu LTS22.04 and has installed PHP 8.1.2-1ubuntu2.10.
The device uses no device template and has no associated data queries or data templates.
But there is one apache error message when try to enable the device:

PHP Fatal error: Uncaught TypeError: Unsupported operand types: string * int in /var/www/cacti/lib/api_device.php:1507\nStack trace:\n#0 /var/www/cacti/host.php(141): api_device_ping_device()\n#1 {main}\n thrown in /var/www/cacti/lib/api_device.php on line 1507, referer: https://xxx/host.php?action=edit&id=125

[Tomypop]

The 500 Internal Server Error is simply a general indication that something's wrong on the server side. Several things could cause this, but it's always on the website server and not an issue with your computer or internet connection. In most cases, this means the server is down.1

[TheWitness]

Yea, 8.1.x is not even technically certified for Cacti 1.2.x, though we have been fixing issues as they come up. So, for people like me that are on the edge, I'm getting literally no 50x errors right now.

I would update both to the pre-release of Cacti 1.2.24, and as many plugins as throw errors. In the 1.2.24, you can receive an Email any time a users experiences a 50x error message, which is nice. I've fixed a few pages already due to that.

[jankovich]

I would like to add a UPS/netman plus to our v1.2.22 Cacti setup.
But adding show me the following error message:
Sorry, we could not process your last action.
Error:
500 Internal Server Error
Reason:
SNMP Information
There must be something unexpected with this UPS. But I can't see what.
The same UPS is working on our old 0.8.8b Cacti.
All other UPS are working on the v1.2.22. Well they all use different network controller.
I can do a snmpgetnext -v2c -c<community> <ip> .1.3 and I receive an answer. Not the typically sysDescr.0, would that be important? But this behavior is not new, it allways worked liked this.

What could I do to go deeper? There is a device debug switch but there seems nothing specific written in any cacti-logfile.
Thanks for any advice.

Hi I have the same problem with different note. it says

2.png (90.77 KiB) Viewed 8934 times

. I looked at cacti.log and this is the log

1.png (8.39 KiB) Viewed 8934 times

. Im new on linux and i've tried so many chmod code for this but nothing happen. please tell me what to do.

Distro : Almalinux 8.5
Cacti : 1.2.23
php : 8.0
mariadb : 10.3.35

[TheWitness]

You should really not be extending this post as your issue is completely different. Start a new post.

[GGA]

Meanwhile I've upgraded to the latest cacti and spine 1.2.24 but that didn't solve it.
So I took a moment to check if cacti ever sends a SNMP-request to the device.
This is maybe something very strange.

Code: Select all

BAD UPS
snmpget DEVICE1 snmpFrameworkMIB.snmpFrameworkMIBObjects.snmpEngine.snmpEngineTime.0
SNMPv2-MIB::sysDescr.0 = STRING: NetMan 100 plus

WORKING UPS
snmpget DEVICE2 snmpFrameworkMIB.snmpFrameworkMIBObjects.snmpEngine.snmpEngineTime.0
SNMP-FRAMEWORK-MIB::snmpEngineTime.0 = INTEGER: 10625657 seconds

[GGA]

Compared, this is what old 0.8.8b Cacti does.

Code: Select all

IP cop.46418 > usv.snmp:  C="xxx" GetRequest(28)  system.sysDescr.0
IP usv.snmp > cop.46418:  C="xxx" GetResponse(45)  system.sysDescr.0="NetMan 100 plus"
IP cop.53446 > usv.snmp:  C="xxx" GetRequest(28)  system.sysUpTime.0
IP usv.snmp > cop.53446:  C="xxx" GetResponse(34)  system.sysUpTime.0=3832488772

So the old cacti asks for sysUpTime while the actual one asks for snmpEngineTime.
Could it be because of the wrong return value? STRING instead of INTEGER?

[TheWitness]

Sorry, but having to lock this one. Please start new threads independently. Apologies.