Software probe won't reconnect after being disconnected for 2 weeks

RIPE Atlas
1

Hi there,

I’m posting this on behalf of a (currently unavailable) friend whose software probe 1008470 won’t reconnect after being down for 2 weeks (this was a planned maintenance on his side).

The logs look like this:

ripe-atlas-1 | /run/ripe-atlas/status/reginit.vol does not exist try new reg
ripe-atlas-1 | Ping works
ripe-atlas-1 | start reg
ripe-atlas-1 | ATLAS registration starting
ripe-atlas-1 | REREG_TIMER 1754726353 expired now is 1756105940
ripe-atlas-1 | REREG_TIMER_EXPIRED go re register REREG_TIMER 1 , now is 1756105940
ripe-atlas-1 | REGHOSTS reg03.atlas.ripe.net 193.0.19.246 2001:67c:2e8:11::c100:13f6 reg04.atlas.ripe.net 193.0.19.247 2001:67c:2e8:11::c100:13f7
ripe-atlas-1 | ssh -p 443 atlas@2001:67c:2e8:11::c100:13f6 INIT
ripe-atlas-1 | 255 REGINIT exit with error
ripe-atlas-1 | condmv: not moving, destination ‘/var/spool/ripe-atlas/data/out/v6addr.txt’ exists
ripe-atlas-1 | condmv: not moving, destination ‘/var/spool/ripe-atlas/data/out/simpleping’ exists

even changing the /etc/ripe-atlas/probe_key* ssh key doesn’t help. Does any of you know a way around it?

Cheers

Denis

2

Hi,

Thanks @dot for the post.
I went a bit further with my troubleshooting. It looks like my probe key has not been deployed on all Ripe’s target servers.

The server which refused my key this morning, now accepts it.

ssh -vv -p 443 -i /etc/ripe-atlas/probe_key -T atlas@2001:67c:2e8:11::c100:13f6 INIT 2>&1 | grep "accepts"
debug1: Server accepts key: /etc/ripe-atlas/probe_key RSA SHA256:Fq6dAwH/PiDJtCWYh/EklbuiCti6vCTTyhykcqCvcjc explicit

But my probe now tries to connect to 193.0.19.25which refuses the connection.

ssh -vv -p 443 -i /etc/ripe-atlas/probe_key -T atlas@193.0.19.25 INIT 
atlas@193.0.19.25: Permission denied (publickey).

Greetings,

Pierre-Yves

3

Hello, the most likely cause is that the probe is using a different key than what’s on record. As the host you can change / update this for SW probes in the UI.

If this doesn’t help then can you share your probe id? It’s best to do this in a ticket to atlas@ripe.net with the pub.key included so we can check.

4

Hello, I have tried several times to delete the two files /etc/ripe-atlas/probe_key*and upload the freshly re-generated pub file on RIPE Atlas - RIPE Network Coordination Centre (using the “Update SSH key” feature) ; the probe kept reporting 255 REGINIT exit with error messages.

Will create a ticket then.