Device status changes from up to down

[Nadou21]

Hi everyone!
Lately I've been having an issue with some of my devices in cacti (cisco switch L2), The state of the device changes every minute from down to recovering, but in fact the device is always up, and it still graphing data.
The devices responds to snmp, but has error in TCP which is weird because it has access and there is nothing blocking it.

Code: Select all

SPINE: Poller[Main Poller] PID[14556] Device[xxx] Hostname[x.x.x.x] ERROR: HOST EVENT: Device is DOWN Message: Device responded to SNMP, TCP: Cannot connect to host 

The problem is that I keep receiving those error messages in my logs, and have some broken graphs of cpu.

This is from debug log:

Code: Select all

SPINE: Poller[Main Poller] PID[16176] Device[] DEBUG: Entering SNMP Ping  
2021.04.06 16:16:04 - SPINE: Poller[Main Poller] PID[16176] Device[] DEBUG: TCP Device Alive, Try Count:1, Time:26.9918 ms  
2021.04.06 16:16:04 - SPINE: Poller[Main Poller] PID[16176] Device[] DEBUG: Entering TCP Ping  
2021.04.06 16:16:04 - SPINE: Poller[Main Poller] PID[16176] DEBUG: The Value of Active Threads is 42 for Device ID   
2021.04.06 16:16:04 - SPINE: Poller[Main Poller] PID[16176] Device[] HT[5] DEBUG: HOST COMPLETE: About to Exit Device Polling Thread Function  
2021.04.06 16:16:04 - SPINE: Poller[Main Poller] PID[16176] DEBUG: The Value of Active Threads is 44 for Device ID   
2021.04.06 16:16:04 - SPINE: Poller[Main Poller] PID[16176] Device[] HT[6] DEBUG: HOST COMPLETE: About to Exit Device Polling Thread Function 

From web interface, the SNMP info is displayed, and ping results is :

Code: Select all

TCP Ping Success (27.89 ms)

Any help to resolve this.
Thanks!
Cacti: v 1.2.16
poller: spine 1.2.16