SNMP Walk & Get Timeout on some device

[bgrunhovd]

Cacti Version: 0.8.8c
Server: Windows 2008 R2
NET-SNMP: Version: 5.5
RRDTool: 1.4
PHP Version: 5.6.3
Memory Limit: 128M
Poller Type: Spine (started with cmd)
Max Threads: 10
Number of PHP Script: 5
Script Timeout value: 180
Max SNMP OID per SNMP get request: 50

When i upgraded my server after i rebuilt it. I notice that some of my devices i couldn't get data to populate for the Interface Statistics in the items and rows. I notice in the basic SNMP information like uptime, hostname, location & contact would populate, but nothing else.

when i run the reload data query it would sit there and think for about 30 seconds and it would bring back nothing or dump me to a "Server Error 500 - Internal server error message". when i go into the log i see the following:
03/19/2015 09:14:44 AM - CMDPHP: Poller[0] WARNING: SNMP Walk Timeout for Host:'10.100.60.141', and OID:'.1.3.6.1.2.1.31.1.1.1.18'
03/19/2015 09:14:14 AM - CMDPHP: Poller[0] WARNING: SNMP Get Timeout for Host:'10.100.60.141', and OID:'.1.3.6.1.2.1.2.1.0'

This is a device that i could have about 1500 items and about 200 rows, but if i take the same device and try to walk it with no interfaces up then it would return data for me.

Looks likes it timing out, but i can't figure out what i'm missing on it.
Hope i have given you enough information.
thanks

[BSOD2600]

does the device implement anti-DDoS or similar to protect itself and thinks the Cacti server is attacking?
try increasing the execution time in php.ini (per windows install guide)?
try increasing snmp timeout for the device in cacti?
try decreasing the max oids per request?
try changing between snmp v1 and v2?

[bgrunhovd]

does the device implement anti-DDoS or similar to protect itself and thinks the Cacti server is attacking?
Nope, when i have no endpoints connected to the device it walks and gets information just fine. Its just when there are a lot of endpoints then it fails when i do the reload data query or Verbose Query. It brings back the "Server Error 500 - Internal Server error"

try increasing the execution time in php.ini (per windows install guide)?
Changed it up to 240 and still have same issue.

try increasing snmp timeout for the device in cacti?
Changed it up to 500000 and still does it. It will only try to walk it for about 20-30 seconds and then errors out.

try decreasing the max oids per request?
Changed it down to 2 and nothing.

try changing between snmp v1 and v2?
Device will only take v2

[bgrunhovd]

It looks like its a FastCGI process that is timing out. Not sure what to do to fix this.

[bgrunhovd]

So after doing some research on the FASTCGI error.
In my IIS7 i had to go into the FASTCgi settings and change the Process Model > Activity Timeout up. By default it was set for 30 seconds so one that timeout hit it would timeout the request. I just changed it up to 600 sec and it seem to pull the data.

Under the clog i still will get a handful of get and walk timeout from different OIDs though.

[BSOD2600]

So after doing some research on the FASTCGI error.
In my IIS7 i had to go into the FASTCgi settings and change the Process Model > Activity Timeout up. By default it was set for 30 seconds so one that timeout hit it would timeout the request. I just changed it up to 600 sec and it seem to pull the data.

http://docs.cacti.net/manual:088:1_inst ... _webserver links to http://www.iis.net/learn/application-fr ... ons-on-iis which does have that max_execution_time 30 sec timeout. Interesting that devices takes longer to respond, but good into/troubleshooting for future Googlers

Under the clog i still will get a handful of get and walk timeout from different OIDs though.

are the OIDs actually resolvable? is the device too busy to handle SNMP requests?