Skip to Content.
Sympa Menu

perfsonar-user - Re: [perfsonar-user] Re: [perf-node-users] Cacti/SNMP collection via the perfSONAR toolkit

Subject: perfSONAR User Q&A and Other Discussion

List archive

Re: [perfsonar-user] Re: [perf-node-users] Cacti/SNMP collection via the perfSONAR toolkit


Chronological Thread 
  • From: Hector Ordorica <>
  • To:
  • Cc: , "" <>, "" <>
  • Subject: Re: [perfsonar-user] Re: [perf-node-users] Cacti/SNMP collection via the perfSONAR toolkit
  • Date: Wed, 12 Mar 2014 12:48:45 -0700

I tried to set it up on one of our perfSonar instances, but found the cacti database to be corrupt (not fully updated I believe) on a fresh install of 3.3.2. I couldn't add a router to monitor. Perhaps I misconfigured it at some point.

I eventually destroyed the database and imported cacti.sql to replace it, and added a router.

I think it's useful for monitoring the perfSonar box itself, but found the default settings for bandwidth monitoring lacking to catch the 20 second flows the throughput tests have. I tend to monitor those links with other software which polls at a much more frequent interval to get a more accurate bandwidth usage.

I don't mind it, but I probably won't be spending more time configuring it to improve it.

Hector Ordorica
Unix & Network Application Support
UC San Diego

On Wed, Feb 12, 2014 at 8:24 AM, Michael Tiernan <> wrote:
On 2/12/14 11:00 AM, Brian Tierney wrote:
We'd like to know if anyone is currently actually using Cacti, which is bundled on the perfSONAR toolkit, for collection of SNMP data.

I/we don't. I feel the machine is too exposed to the public net for my taste.
If we wanted to use it, I'd set it up on a machine I have hiding behind firewalls and using other network resources to protect.

(Others, more skilled/experienced than I may be willing/able to risk this but I am erring on the side of caution and assuming that the black-hatted "them" are more wiley than I am. To this end, I try to keep our exposure to a minimum.)

--
  <<  MCT>>    Michael C Tiernan +1 (617) 324-9173
  MIT - Laboratory for Nuclear Science - http://www.lns.mit.edu
  High Perf Research Computing Facility at The Bates Linear Accelerator
    Please avoid sending me MS-Word or MS-PowerPoint attachments.
    See http://www.gnu.org/philosophy/no-word-attachments.html






Archive powered by MHonArc 2.6.16.

Top of Page