Mine eventually worked.
I should have started a new thread. Though I still can.
Mine eventually worked.
I should have started a new thread. Though I still can.
Thanks for the reports! We are aware of degraded issuance, and have opened an incident on letsencrypt.status.io while we investigate. Sorry for not doing so sooner.
Hi, Iād like to report another instance of the same problem.
We get the same error message:
{
"type": "urn:acme:error:serverInternal",
"detail": "Error creating new cert",
"status": 500
}
from the service
POST https://acme-v01.api.letsencrypt.org/acme/new-cert
We have been using Lets Encrypt to provision certs for domains for a while without such problems. The problems started appearing when we tried to issue certs for some new domains:
training5.evidentiae.ca
training6.evidentiae.ca
training7.evidentiae.ca
training8.evidentiae.ca
training9.evidentiae.ca
api-training5.evidentiae.ca
api-training6.evidentiae.ca
api-training7.evidentiae.ca
api-training8.evidentiae.ca
api-training9.evidentiae.ca
upload-training5.evidentiae.ca
upload-training6.evidentiae.ca
upload-training7.evidentiae.ca
upload-training8.evidentiae.ca
upload-training9.evidentiae.ca
Weāre using simp_le
, and it aborts on receiving the error from the server. I donāt know on exactly which subset of the above domains the error occurs.
When we removed these domains from the list of domains that we provision using simp_le
, the problem disappeared.
I can provide more detailed logs if you want.
Thanks,
David
We have noticed that the problem happens when adding another new set of domains (named something else than ātraining-ā), and actually the problem might simply be triggered when simp_le
notices that some domain has changed and starts provisioning certificates for our entire set of domains. So the problematic request could very well trying to request a certificate for one of our pre-existing domains. What Iām really trying to say is we donāt know which domain that triggers the problem. But in any case we are still seeing this internal error from the Lets Encrypt server today.
@jsha there are also reports of continied 500s in another thread, which might be related.
From @mnordhoff in production:
and with staging from @larryboymi:
EDIT: I missed that the second reproducer hadn't tried again since last weekend so this one may not be useful (unless it's a totally different issue).
Weāve discovered that the error we see happens when we pass the limit of 100 names per certificate. Sorry for the noise.
Precisely what error message do you get? If it's "error creating new cert", it's still a bug that Boulder isn't saying something more specific.
Yes, the fact that we got a 500 Internal Error is what made us confused in the first place. The response body was:
{
"type": "urn:acme:error:serverInternal",
"detail": "Error creating new cert",
"status": 500
}
David
Closing the loop, hereās the corresponding Boulder issue: https://github.com/letsencrypt/boulder/issues/3632. Thanks for all who helped out in this thread!
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.