In response to Zate's question:
 
I scan ~9000 devices in a monthly cycle. I run 1 Nessus scanner on 1.0 GHz PIII under Debian Linux
 
I deliver monthly reports relating to suspected problems and some next working day reports about critical problems to relevant departmental contacts (for multiple departments/network organisational units).
 
I update Nessus signatures and Nessus if necessary monthly then restart Nessus server. After adjusting the plugins config I taking a copy or the relevant .nessusrc file that I will be using for batch scans.
 
I use my own system to automate scanning written in Perl calling FreeTDS to access remote MS SQL databases. The scripts are run by cron and download the next batch of addresses to scan, run the scan then upload the results to the SQL database (I parse the NBE format output using Perl). There are some bells-and-whistles such as reporting, locking, logging and stopping scans that have not completed on time (just in case) of course.
 
A table in my SQL database is updated regularly with ARP cache data from all the switches so that I can focus on devices that have been live recently on the network. I do not rely on response to ping because many systems do not respond now (XP firewall). Completely blind scanning would waste a huge amount of time. Only scanning registered IP addresses could miss rogue devices. We have an IP device registration database in which registered devices are classified by type etc. I do not only scan for registered devices, however, the database allows me to avoid routinely scanning network infrastructure devices. I can also decide whether to scan a device depending on how recently I scanned it, whether to exclude it for some reason, etc. using SQL data. New results for host are inserted after any old results are deleted. Scan data is purged from the database if and when it reaches month old.
 
I have written a VB program that allows me to select Nessus results for a department or group (or department unknown) from the database (this is possible because of our IP registration database). I can then select records on the basis of risk classification, OS version, IP number, subnet, port, CVE etc. I have a button to make selection of the most concerning records easy. I can then send an email with attached report addressed by default to the correct contact. This stage is semi-automatic so I can review the data and make judgements and include comments in the email (I don't think Nessus can be used well without some experience because it might give advice like "disable this..." without even knowing the context). I use a SQL data driven Excel spreadsheet to give me an overview report and show which departments/groups need to be sent a report this month. When records are emailed out I can flag in the DB which ones have been reported so I don't repeat myself. Departmental contacts can inform me if they have concluded that some IP address/Plugin number combination is providing false-positives that they don't want to see again; I can feed that back into my DB to modify how data is selected for subsequent reports. 
 
I have some daily scheduled SQL routines that email me about new items of interest in the data (such as banners saying "drug of choice"), FTP servers running on unusual port number etc.
 
(As you see I have done a lot of stuff myself. I think I started doing some of it before off-the-shelf solutions began to appear. If I was starting again I would do some things differently i.e. make even more use of SQL server. Nevertheless, I doubt that anything off-the-shelf would be as customised or customisable (by me) for my enterprise as what I now have in place.)  
 
--
Carl Nelson
Distributed Systems Support Section, Computer Centre, University of Leicester, Leicester, LE1 7RH, U.K.
Tel: +44 (0)116 252 2060, Fax: +44 (0)116 252 5027
 


From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Zate Berg
Sent: 09 November 2006 14:14
To: [email protected]
Subject: Nessus in the Enterprise

Good Morning All,

I was wondering if anyone could contact me off the list to discuss how they have Nessus setup and deployed in a large network.  I am not finding much information on things like reporting and a centralized web interface.

Mainly looking for info such as

* what you run it on,
* how many scanners you use,
* how you manage user access to the scanners,
* do you use a central Web console of some kind? (does a full featured one exist?)
* How do you store your reports?

Thanks :)

--
Zate
_______________________________________________
Nessus mailing list
[email protected]
http://mail.nessus.org/mailman/listinfo/nessus

Reply via email to