DOCSIS Stats

BSOD2600 · Posted: Wed Oct 04, 2006 2:18 pm Post subject:

You didn't run the script through the cacti script server

You'll also want to specify the various arguments the script wants. Look in your poller cache what Cacti already feeds it.

3AHO3A · Joined: 04 Oct 2006 Posts: 30 Location: Somewhere in EU

3AHO3A · Joined: 04 Oct 2006 Posts: 30 Location: Somewhere in EU

I did itt, graphs contains data now ;]
Thanx for support
Cacti Rulez

Ierlandfan · Joined: 24 Jun 2007 Posts: 3

I am using Cacti 0.6I and having the same problem.

Using the cmd php script_server.php everything is ok.
SNMPwalk ok
But no graphs...

But the log keeps telling me:

06/24/2007 08:40:13 AM - THOLD: Poller[0] Checking Thresholds
06/24/2007 08:40:10 AM - SYSTEM STATS: Time:6.6728 Method:cmd.php Processes:1 Threads:N/A Hosts:4 HostsPerProcess:4 DataSources:20 RRDsProcessed:14
06/24/2007 08:40:09 AM - CMDPHP: Poller[0] Host[2] DS[22] WARNING: Result from SNMP not valid. Partial Result:

In the beginning the error was this:

06/24/2007 07:50:13 AM - THOLD: Poller[0] Checking Thresholds
06/24/2007 07:50:10 AM - SYSTEM STATS: Time:5.5184 Method:cactid Processes:1 Threads:1 Hosts:4 HostsPerProcess:4 DataSources:20 RRDsProcessed:14
06/24/2007 07:50:09 AM - CACTID: Poller[0] Host[2] DS[22] WARNING: Result from SNMP not valid. Partial Result: ...
06/24/2007 07:50:09 AM - CACTID: Poller[0] Host[2] ERROR: Problems parsing Multi SNMP OID! (oid: )

I changed a lot, imported it as xml, removed it, copy, paste script import, added spaces between the <brackets> in data input method, set catid to cm.php but it didn't do the trick.
By the way view poller cache? It is in SNMP cache?

Anyone know what to check? I am little lost here.

BSOD2600 · Posted: Sun Jun 24, 2007 11:18 am Post subject:

1) DS[22] is this DOCSIS script?
2) snmpwalk .1.3.6.1.2.1.10.127.1 to verify your modem responds to DOCSIS related queries.
3) try using cmd.php instead and see if that makes a difference with the parsing Multi SNMP OID error.

Ierlandfan · Joined: 24 Jun 2007 Posts: 3

1. DS is docsis script
2. SNMPwalk is ok. Lot of response
3.cmd.php gives me the following:

06/24/2007 12:05:13 PM - THOLD: Poller[0] Checking Thresholds
06/24/2007 12:05:09 PM - SYSTEM STATS: Time:5.7824 Method:cmd.php Processes:1 Threads:N/A Hosts:4 HostsPerProcess:4 DataSources:20 RRDsProcessed:14
06/24/2007 12:05:09 PM - CMDPHP: Poller[0] Host[2] DS[22] WARNING: Result from SNMP not valid. Partial Result:

(by the way, my scripts dir is under /var/www/html/scripts not under /var/www/html/cati/scripts)

script server cmd line: php script_server.php

/var/www/html/cacti/scripts/ss_docsis_stats.php ss_docsis_stats 192.168.100.1 public 1 161 500

docsIfDownChannelPower:36 docsIfSigQSignalNoise:376 docsIfSigQMicroreflections:39 docsIfCmRangingTimeout:5 docsIfCmStatusTxPower:438 docsIfCmStatusResets:0 docsIfCmStatusLostSyncs:0 docsIfCmStatusT1Timeouts:0 docsIfCmStatusT2Timeouts:0 docsIfCmStatusT3Timeouts:1 docsIfCmStatusT4Timeouts:0

maybe it was a permission thing, I got a permission denied, so I chmodded it, who knows..) testing as we speak.

Ierlandfan · Joined: 24 Jun 2007 Posts: 3

Right, it works.

I deleted everything. (including host)
Uploaded everything again.
Made the host (without settings anything for example *not* setting generic-snmp-enabled-host) set it to none
make graph

Turned on debugging on log file. Saw the lack of spaces (Although I changed them and saved it) so a little confused i did:

Changed input query to %<>% and saved
Error (expected) Showed up in log (which is what I wanted)
Changed the % to spaces.
Cleared the poller,
Deleted the rrd files
Log ok!

and added graphing
And for some reason it works now!

Thanx, the debugging NaN topic bought me the ideas

roygrubb · Joined: 19 Sep 2007 Posts: 8

Most of our modems are Surfboard 5100s and it seems as if everything imported fine. I created the graphs via the device and verified that graph management is showing the graph meaning it created an .rrd just fine. The only thing is all fields are reporting "nan" as no data is being polled. I turned on debug and everything looks to be fine there as well. I would love to be able to get this information graphing in Cacti.

roygrubb · Joined: 19 Sep 2007 Posts: 8

our php script was in the wrong place as we're running Ubuntoo
Works great! thank you so much

cableguy34 · Joined: 18 Oct 2007 Posts: 1

I am looking for Cable modem templetes for Cacti with SNMP walk or using SNMP. I also would like to monitor Rx, Tx, LostSyncs, Resets, T1, T2, T3, T4, For some reason i am having a hard time building my own mibs. so if anyone can assist, i would really appreciate it.

thanks everyone

BSOD2600 · Posted: Thu Oct 18, 2007 3:41 pm Post subject:

Ummm, did you even try the template in this post?

jt72 · Joined: 14 Jul 2008 Posts: 5 Location: Cork, Ireland

First of all, thanks for this script! I'm using it with Cacti 0.8.7 and having pretty good success. It's also a very useful example for gathering stats from DOCSIS devices.

I'm wondering if it's useful getting the RangingTimeout though, because it's not reallya counter, more of a setting. It's the time a CM will wait before incrementing the T3 Timeout counter if I remember correctly...

jt72 · Joined: 14 Jul 2008 Posts: 5 Location: Cork, Ireland

BSOD2600 · Posted: Tue Jul 15, 2008 6:30 pm Post subject:

One should only need to create ONE data source with the DOCSIS, since as you've seen, it executes all of the queries in one go. These are all stored in the same rrd file which the graphs both pull from.

Hmm, looks like the script should probably be updates to return NaN if no data was returned...

jt72 · Joined: 14 Jul 2008 Posts: 5 Location: Cork, Ireland

I have a couple of modems which are reporting downChannelPower of > 80 (so +8.0 dbMv). I think as a result the values are being dropped because they're out of the accepted input range (80 to -80?).

The range for DOCSIS modems is about +15 to -15, and some would work in a wider range. Does this mean we should update the graph template to use +200 to -200 as the valid range?

For example the first 2 CMs will NOT have the downChannelPower graphed (87 and 85 respectively) while the other one will.