-
-
Notifications
You must be signed in to change notification settings - Fork 69
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
0 ROAs from ripe or apnic -- how to troubleshoot? #246
Comments
APNIC and RIPE are the two regions that use RRDP, so it probably is related to that. You can verify by running Routinator with the |
thanks @partim I will give that a try. Is there anything I need add / install to enable RRDP? In my config file I have 'disable-rrdp = false'. I run routinator with |
i updated my config file to log to local file and level INFO, and also disabled rrdp, and i see in the log all the rsync messages seemingly complete. and indeed, on /metrics I now see ROAs for all the RIRs:
I will try re-enabling RRDP and see if any logs point direction to why it fails. |
well, now i'm thoroughly confused, as it seems to have everything working, at least for now (maybe some are cached from the prior rsync run?). info from metrics after restarting with RRDP enabled again:
for ca.rg.net RRDP failure, this was seen in the log:
ill go through the same process on my other instance and see if it recovers for the time being. i'll check it again tomorrow and see if it still reports data from all the RIRs. EDIT: the above output was when I ran
this was within minutes of it seeming to succeed in the prior output -- and the metrics for rrdp status all report -1 again:
not sure one has any cause for the other, will do a few more restarts and see if its consistent. additional edit: after several restarts on both instances I have, if I do Thanks, Will |
Interesting! Thanks for figuring that out! |
The reason why you ended up with 0 ROAs for the RRDP-capable regions is that while Routinator will fail over to rsync if RRDP doesn’t work at all, it will not if it has what appears to be a valid local copy of an RRDP repository. This is so that it won’t fail over for intermittent RRDP failures that would fix itself on the next run. In this case it will just use the existing local copy. I suspect that in your case, it did exactly that but since it didn’t update anymore, that copy slowly expired, eventually leading to 0 ROAs. PR #250 fixes the failure to run RRDP with the I believe we never discovered the issue since hardly anyone actually uses that option – in most cases the init system seems to want to do its own forking. For the same reason, I am tempted to not do an urgent release for the fix. |
Let’s move this to a separate comment for posterity: My impression is that the |
Thanks @partim - what is the most common method to run routinator? via systemd? I saw the example unit file, but myself am not familiar with how to incorporate -- if that is the best / preferred way to run, i'll hit up my colleagues that are more systems oriented to help me figure it out with how I currently have routinator installed (also likely not ideal...right now just in my user home directory on the instance i'm running) Thanks again! |
On Linux, I would indeed suggest to use systemd. There is two service unit files, the minimal one should work on any version of systemd, the regular one needs a newer and we had issues with it leading to segfaults on some systems. The service assumes a config file similar to etc/routinator.conf.system-service. If you are using Debian-based systems, you can also build a Debian package via cargo deb, but I am not quite sure how well that works currently. It definitely doesn’t do the Alternatively, Docker via Docker Hub seems to be very popular. |
yea i think longer term we'll probably look to deploy as a docker container, but in the interim i'll work on getting the unit file working. Thanks, Will |
Hello,
I have two installations of routinator (just upgraded, but this issue predates the upgrade I'm almost certain). somewhere along the way, it seems my instances are no longer able to get ROAs from one or more regions. One instance reports 0 ROAs from apnic and ripe, and another 0 from just apnic. example output from the /metrics:
i can't seem to find any logs to see why one or more seem to be failing. Looking at other metrics being reported it looks like rsync exit codes 0 for all of the repositories (and it seems RRDP is failing with code -1 for the ones it is attempting).
Any feedback as to how to troubleshoot or where to look would be appreciated!
Thanks!
The text was updated successfully, but these errors were encountered: