RIPE Recommendation on Operational Contacts

1. Purpose

Since there is no managed RIPE backbone with central operations staff, operational problems have to be solved in a distributed manner. For this to work, network configuration information, network status information and operational contact information must be available to the operations staff of all RIPE members. With this information, the operations staff of a RIPE member can pinpoint the causes of a problem and contact the appropriate operations staff of another member quickly. Without this information, locating problems would either take people and time or be impossible.

The purpose of this recommendation is to give guidance to the RIPE member organisations on how to exchange contact information and network status information in a way which is easy to implement and which provides enough redundancy to function when network connectivity is not guaranteed. This recommendation only covers a simple, rather informal notification scheme. Recommendations about formal trouble ticketing systems (which is under investigation by the IETF User Connectivity Problems Working Group) are beyond the scope of this recommendation.

2. Operational Mailing List

A mailing list called ripe-op should be used to transmit operational information which is of interest to the operational staff of all RIPE members.

Normally mailing lists are maintained and expanded at a specific host. Messages intended for the list are sent to that host from where a copy of the message is sent to each member of the list. The ripe-op list is going to be needed most urgently in case of operational failures. During such failures some places may be isolated. In this case the list would not be useful if it was expanded at only one place. Therefore the list will be expanded locally as much as possible in order to gain as much redundancy as possible. So every RIPE member organisation is urged to provide an expansion point.

The list of mail addresses itself will be made available in the RIPE Information Server as file ripe-op-list. This list will be installed and kept up to date at all expansion points. The list of expansion points (hosts) will be made available in the RIPE Information Server as file ripe-op hosts. This list will contain one line per host with two strings separated by white space. The first string will be the fully qualified domain name of the host. The second string will be the mail address of the maintainer of the local expansion point. This mail address will be notified every time ripe-op-list changes.

Changes to ripe-op-list should be sent to mailto:ripe-op-request@ripe.net.

All messages to the ripe-op list will be archived on the RIPE Information Server. The last few messages can be viewed by telenetting to TCP port 7473 at ftp.ripe.net.

3. Messages on the Operational Mailing List

Messages to ripe-op should be short information about events affecting network operations including:

  • link failures
  • router failures
  • scheduled outages
  • degraded performance
  • excessive traffic
  • security-related incidents

The messages should contain at least:

 

  • a time stamp of the event concerned (not the message)
  • a short description of the problem
  • likely areas affected
  • an address to contact for more information (if not the sender)
  • an estimated time to fix the problem
  • actions taken to fix the problem
  • an indication whether the message is generated manually or automatically

A suggested format for messages is defined in section 4. Care should be taken to also report to the list when the problem has been fixed using a second type of message.

Discussions on the ripe-op list are discouraged.

4. Suggested Message Format

The example messages below are included as a suggestion only. They do not imply a recommendation for a (specific) formal ticketing system.

_____________________________________________________________________________
| |
|From: operator@ripemem.net |
|To: ripe-op@ripemem.net |
|Subject: TICKET ripemem #4711 OPEN |
| |
| |
|Ticket Number: 4711 Ticket Status: open |
|Ticket Type: unscheduled Ticket Source: ripe-member |
|Ticket Scope: link Site/Line: ripemem, otherripe |
|Ticket Owner: operator@ripemem.net Problem Fixer: |
| |
|Ticket Opened: 910516 11:40 UTC Problem Started: 910516 11:10 UTC |
| Closed: Ended: |
| |
| |
|Problem Description: |
| |
| Link between routers router1.ripemem.net and |
| gateway3.otherripe.net is down. |
| |
|Affected: |
| Traffic between ripemem and otherripe is re-routed |
| via nice-net. Performance may suffer due to limited |
| bandwidth |
| |
|Actions: |
| Coordinated with operator@otherripe.net. |
| PTT in ripemem-land informed (ticket #1234). |
| PTT in otherripe-state being informed. |
| |
|Time to fix: |
| no estimate yet |
|___________________________________________________________________________|

_____________________________________________________________________________
| |
|From: operator@otherripe.net |
|To: ripe-op@otherripe.net |
|Subject: TICKET ripemem #4711 CONTINUE |
| |
| |
|Ticket Number: 4711 Ticket Status: open |
|Ticket Type: unscheduled Ticket Source: ripemem |
|Ticket Scope: link Site/Line: ripemem, otherripe |
|Ticket Owner: operator@ripemem.net Problem Fixer: |
| |
|Ticket Opened: 910516 11:40 UTC Problem Started: 910516 11:10 UTC |
| Closed: Ended: |
| |
| |
|Actions: |
| PTT in otherripe-state informed at 1205 UTC (ticket #5678) |
| |
|Time to fix: |
| no estimate yet |
| |
|___________________________________________________________________________|


_____________________________________________________________________________
| |
|From: operator@ripemem.net |
|To: ripe-op@ripemem.net |
|Subject: TICKET ripemem #4711 CLOSE |
| |
| |
|Ticket Number: 4711 Ticket Status: close |
|Ticket Type: unscheduled Ticket Source: ripemem |
|Ticket Scope: link Site/Line: ripemem, otherripe |
|Ticket Owner: operator@ripemem.net Problem Fixer: operator@ripemem.net |
| |
|Ticket Opened: 910516 11:40 UTC Problem Started: 910516 11:10 UTC |
| Closed: 910516 14:30 UTC Ended: 910516 14:25 UTC |
| |
| |
|Fix: |
| PTTs found multiplexer problem. |
| |
|___________________________________________________________________________|