From Best Practices

Jump to: navigation, search

Welcome to Best Practices. On this site, you'll find the collected wisdom of our contributors (who are probably gacking behind a server somewhere at my phrasing it that way :-) on the most important aspects of internetworking operations and the best approaches to take to these services and tasks. If you'd like to contribute from your experience as well, please start out by taking a look at our Community Page for some hints and philosophy.

This is a wiki, so the material you read has been contributed and edited by many people; you will likely find it important to check the history page and the User Talk pages for the listed contributors to properly adjust your confidence factor in the material you find here.

As with most reference works on complicated topics, please remember that this site should not be construed as providing medical, legal, engineering, or any other form of advice the commercial provision of which in your jurisdiction requires licensure. Make sure you understand what you're doing before you do it: if you break something, you get to keep both pieces.

And, as with most wikis, this site will always be a work in progress. You might find it useful to bookmark the Recent Changes page to keep an eye on the evolution of the whole site, or add specific topic pages to your Watchlist.

Gratuitous plug: Anyone responsible for managing and designing internetworks needs to be regularly reading the RISKS Digest. Though not all the items in each digest relate directly to network management, almost all of them have some aspect which it would be useful for network managers and designers to keep in mind.

[edit] Best Practices Philosophy

What is a "best practice" -- as distinguished from merely a good idea?

This is a somewhat elusive concept, but people have made attempts to approach the definition in various ways. To begin with, this section will point to some of these attempts as source documents and over time we will develop our own definition to guide the material that appears here. While the rest of this wiki will concern itself specifically with the best practices of Internet Engineering, this section will sometimes touch on non-technical and non-Internet aspects of best practices where those aspects provide useful guidance to both contributors and to those attempting to implement best practices on their networks.

Here are some thoughts of Paul Ferguson on the Deja Vu issue.

Randy Bush and Dave Meyer are the co-authors of RFC 3439 which discusses Internet architectural guidelines and philosophy. Some people may question why there is the need for a wiki when much of this material is already published elsewhere, especially in RFCs like this one. The main reason is that the wiki allows us to pull all this material together in one place. By doing so, we can make the source documents more widely known and more accessible to people. At the same time we have the opportunity to update best practices in the light of current knowledge.

[edit] Topics

This is the table of contents for Best Practices. New topic sections and links should go here.

[edit] Internetworking

This section is about the principles and practices of moving packets from one place to another, mostly without respect to what they contain. Much reference information is available on this topic at Wikipedia.

[edit] Physical Layer

Circuits | Route diversity | Dealing with vendors | Metro Area Networking | Wide Area Networking

[edit] Internetwork Layer

ICMP: Should you respond to ping?
IPv4:
- Addressing
  - Address allocation | Regional registries | Drop RFC 1918 packets?
- Routing
  - Protocols:
    - Interior Routing Protocols: OSPF | IS-IS | EIGRP | RIP
    - BGP
  - Routing blacklists
IPv6:
- Addressing
  - Address allocation | Regional registries
- Routing
  - Routing Protocols: OSPFv3 | IS-IS | EIGRP | RIPng| BGP
- Connecting to IPv4
  - 6to4 | Teredo | ISATAP | Tunnel Brokers
TCP
Upper-layer Protocols

[edit] Normal Operations & Network Management

Network Operations
- Running an effective NOC
- Never Shed a Tier
SNMP
- Supersize me: Dealing with SNMP collection in large networks.
- SNMP Tools
Syslog
- Events are cowards, they attack in force: How to cope with insane log volumes
- Syslog Tools
Netflow
- Traffic is light and it's still 9 million flows a day: Aggregation and Reporting.
- Netflow Tools
Authentication, Authorization, and Accounting
- TACACS
- Radius

[edit] Outages, Events & Attacks

Types
Identification
Mitigation
Prevention

[edit] Network Services and Applications

Topic in this section cover higher-layer services, how to run them most effectively, and how to avoid abuse.

[edit] Network Time Protocol (NTP)

Intro

[edit] Domain Name System (DNS)

Intro | DNS Blacklists | Unusual Domain Names

[edit] E-mail and Spam <not-a-TM>

The MailScanner people have their own Best Practice page on anti-spam and related mail processing; unsurprisingly, they like MailScanner for this work.

[edit] HTTP, Web Services, etc

[edit] Building for service redundancy

High availability technologies: HSRP | Local Server Load Balancing | Global Server Load Balancing

[edit] Communications

This section is for information, references, and tips related to communicating with the other people from whom you'll need help to fix any networking problem covering equipment or territory you're not in control of. Much of this section will likely be tips and case studies.

Voice Communications

Telephone: PSTN | Cellular | Internet Telephony
Radio: FRS | GMRS | Business Band | Nextel
Other: VoIP

Text Communications

Pagers: Local Paging | Skytel | Iridium | BlackBerry
Network: IRC
- Instant Messaging: AIM | Jabber

[edit] Specific Products, Gotchas and Caveats

Please see the site disclaimer concerning references to specific products by name.

[edit] New Material

New material from contributors that does not have an obvious place to reside in the index above can be linked from here.

Tools and Such

Dymo label makers, for marking cаbles and connectors

Main Page