I’m using a wildcard cert from letsencrypt. Currently there is only one way how to verify that you hold the domain you are requesting cert for: creating TXT record in that domain. You can do it by hand when asked by certbot but you don’t want to do this each 90 days. If you are running your own DNS servers with PowerDNS like me there’s an elegant solution: RFC2136. This allows you to update your zones without writing config files and restarting anything.
Let’s see how to do this:
First check in /etc/powerdns/pdns.conf that you have DNS updates allowed and where from. You can also add another host if you are are requesting your certs from elsewhere. This is default:
################################# # allow-dnsupdate-from A global setting to allow DNS updates from these IP ranges. # # allow-dnsupdate-from=127.0.0.0/8,::1
Next, create a hook script with following content and make it executable:
$ cat <<\EOF > /opt/letsencrypt-dns-hook.sh #!/bin/bash CREATE_DOMAIN="_acme-challenge.$CERTBOT_DOMAIN" echo " server 127.0.0.1 update delete $CREATE_DOMAIN TXT update add $CREATE_DOMAIN 60 TXT $CERTBOT_VALIDATION send " | nsupdate sleep 10 EOF $ chmod +x /opt/letsencrypt-dns-hook.sh
This script takes variables passed in by certbot and creates a nsupdate request to DNS which is then executed (make sure you have nsupdate installed). I also added 10 second delay to allow for the change to propagate to my secondary DNS server.
Then it’s only a matter of using this script with certbot like this:
# /opt/certbot-auto certonly --manual --preferred-challenges=dns --manual-auth-hook /opt/letsencrypt-dns-hook.sh -d *.danman.eu Saving debug log to /var/log/letsencrypt/letsencrypt.log Plugins selected: Authenticator manual, Installer None Cert is due for renewal, auto-renewing... Renewing an existing certificate Performing the following challenges: dns-01 challenge for danman.eu - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - NOTE: The IP of this machine will be publicly logged as having requested this certificate. If you're running certbot in manual mode on a machine that is not your server, please ensure you're okay with that. Are you OK with your IP being logged? - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - (Y)es/(N)o: y Waiting for verification... Cleaning up challenges IMPORTANT NOTES: - Congratulations! Your certificate and chain have been saved at: /etc/letsencrypt/live/danman.eu/fullchain.pem Your key file has been saved at: /etc/letsencrypt/live/danman.eu/privkey.pem Your cert will expire on 2019-01-17. To obtain a new or tweaked version of this certificate in the future, simply run certbot-auto again. To non-interactively renew *all* of your certificates, run "certbot-auto renew" - If you like Certbot, please consider supporting our work by: Donating to ISRG / Let's Encrypt: https://letsencrypt.org/donate Donating to EFF: https://eff.org/donate-le
I hope this post will help someone. Feel free to comment and share.
9 thoughts on “Automatic letsencrypt wildcard cert renewal with nsupdate”
That post really helped me, thy a lot
I just added a ‘pdns_control notify’ a right before the ‘sleep 10’ in order to force/speed up the AXFR to my slaves
Nice tip, thanks!
cool tip 🙂 big thx
Thanks a lot, I thought I had to code this myself, but here you are, saving the day!
Works also brilliantly with bind9 (although I use the shared key authentication from behind a firewall).
Thanks for the information. It works but there are two problems I cannot seem to solve.
Once the script is used, the pdns servers serials get out of sync with which ever domain name is being used. I watched the logs and noticed as soon as I used the script, updates were happening every 60 seconds and never completing with an error that slaves have a newer serial number than the master.
The only way I could clear this was by editing the SOA directly in the database. This could be because I tested a few times in a row.
Second, your command includes manual options. When I try testing this later, it seems that auto renew would not work because the initial command was manual.
I’m not 100% about the second comment because I’ve yet to get past the first problem.
Finally, to prevent the pdns server constantly trying to update that domain, maybe the script should also remove the TXT entry once its done its job?
Something seems to get out of sync. I removed the TXT record from the pdns server yet the slaves are still looking for an update every 60 seconds. I checked to be 100% sure and that record does not show on the master or the slaves. I looked directly in the DB and using the gui.
You should send the update to your master servers, not slaves.
If you are asking if the cert is being generated on the master, yes it is. The slaves are set to check for updates every 60 seconds.
I found out why there is an update every 60 seconds. Someone set up a cron task to force an update to all the slaves by using ‘pdnsutil increase-serial domain.com’ every 60 seconds.