Hi James,

- Recent RRCs have been specified as:

Dell PowerEdge 1750 (1U server) 1 GByte RAM 3 x 73 GByte Disk (configured as RAID5) 2.4 GHz Xeon CPU

Please send me a spec from Dell's web site that's intelligible for MSK-IX folk (should also give us an estimate on price excl BTW). Cheers Matthew

-----Original Message----- From: ris-int-admin@ripe.net [mailto:ris-int-admin@ripe.net] On Behalf Of James Aldridge Sent: 28 July 2004 16:21 To: ris-int@ripe.net Subject: [ris-int] Draft minutes from Monday's RIS meeting (apologies for the delay)

RIS Meeting - 26 July 2004

Present: Matthew, Shane, Arife, James, Bruce, Ruud, Henk, Vasco

Scribe: James

Agenda: - RRC Status - RRC Moves/New RRCs - myASn - AOB

RRC Status

Q. Are there any existing tools to monitor BGP sessions? A. Not that are directly relevent to the RIS. The current tools compare BGP sessions against the registered peering policy (the AS3333 aut-num object).

What to monitor? - percentage of peers up/down? - peer down for a "long time"?

Should there be automatic or manual notification of peers when a session is down?

Who to contact at peers? - We mostly use the IRR (RIPE, etc. databases) - encourage peers to keep their own information up to date - Current "addpeer" script records contact information in a log file - need to keep our own information up to date - useful for historical data (e.g. who was the responsible person at the peer when the peering was first established, etc.)

Action: monitor peering sessions in Nagios (OPS)

peering sessions which are neither up nor administratively disabled should generate alerts.

Action: document procedure for handling peering sessions which are down. (SED)

1. Nagios notifies SED, SED contacts peer 2. After two weeks (?), session is administrively disabled ("neighbor ... shutdown"), (peer is contacted?) 3. After a peering session is down for a further period with no response from the peer then the peering is removed from the RRC ("no neighbor ...")

RRC Moves/New RRCs

* RRC08 (MAE West) has very few peers and more networks are pulling out of MAE West. We need to investigate alternative locations for a route collector on the US West Coast

- PAIX - Seattle - LAIIX (Telehouse) - ...

Action: compare member lists of each West Coast exchange point (Matthew)

How to move RRC? Simply get MCI to ship box to new location (or wait for NCC staff to be in the vicinity of San Jose).

* RRC06 (DIX-IE (NSPIXP2), Japan) - Not getting any new peerings.

Action: Contact George Michaelson (APNIC). Henk and Shane will be at IETF and can meet up with him there.

* RRC13 (new RRC in Moscow)

- Matthew has been in touch with RIPN (Russian Institute for Public Networks).

- To simplify customs issues, it may be better for the host organisation to order machine to our specification (see below) directly from Dell in Russia. We would want to see the quote before agreeing to the purchase.

- We need to know the IP configuration for the box so we can configure and ship the installation CD once the box is ordered.

- Recent RRCs have been specified as:

Dell PowerEdge 1750 (1U server) 1 GByte RAM 3 x 73 GByte Disk (configured as RAID5) 2.4 GHz Xeon CPU

myASn

Meeting with Alexis last week didn't happen (Alexis out of town?)

Action: contact Alexis to set up meeting (Arife)

Matthew's slide "On the road to myASn v.1": - Automate account creation - bring together AS3333/K-Root/RIS peering setup and myASn account creation - For each peering confirmation, the peer should be made aware of the other RIPE NCC peering opportunities, etc. where appropriate.

Action: Check with Dave Knight what, if anything, is mentioned about other NCC services when K-Root peering is established (Bruce)

- Would be easier if all NCC peerings were handled by the same group. - K-Root peerings are more restrictive than RIS peerings.

AOB

RIPE49 (Manchester)

RIS Update presentation myASn Tutorial

Action: Speak to Shane and get back to Matthew et al. (Arife) [Shane had already left the meeting at this point]

Inventory

Ruud asks for Dell Service Tags for each RRC to add to the inventory.

Ops Interaction

Reminder to use <ops@ripe.net> as first point of contact instead of sending email to Bruce/James directly.