I don't think this is an issue with this terraform provider, but I opened a ticket here first because it is easily reproducible with a simple terraform config. Mainly wanted to drop it here to see if anyone else has run into this issue.
Given a sufficiently large set of pass secrets to refresh, we get failed to decrypt errors on plans. This only seems to happen on > 2.1 gnupg. We have tried several version of terraform and gnupg, but our testing seems to point to gnupg > 2.1.
Here is some terraform that will replicate the behavior.
provider "pass" {
refresh_store = false
}
data "pass_password" "dummy" {
path = "dummy/dummy-${count.index + 1}"
count = "100"
}
We consisitently get something like this:
Error: Error refreshing state: 1 error(s) occurred:
* data.pass_password.dummy: 23 error(s) occurred:
* data.pass_password.dummy[46]: data.pass_password.dummy.46: failed to read password at dummy/dummy-47: Failed to decrypt
* data.pass_password.dummy[12]: data.pass_password.dummy.12: failed to read password at dummy/dummy-13: Failed to decrypt
<redacted more errors>
Software versions:
gpg (GnuPG) 2.2.4
Terraform v0.11.7
terraform-provider-pass 1.0.1
I don't think this is an issue with this terraform provider, but I opened a ticket here first because it is easily reproducible with a simple terraform config. Mainly wanted to drop it here to see if anyone else has run into this issue.
Given a sufficiently large set of pass secrets to refresh, we get
failed to decrypterrors on plans. This only seems to happen on > 2.1 gnupg. We have tried several version of terraform and gnupg, but our testing seems to point to gnupg > 2.1.Here is some terraform that will replicate the behavior.
We consisitently get something like this:
Software versions: