You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As a platform engineer, I want the Vultr OpenTofu provider pinned to v2.29.0, so that plan CI doesn't crash on vultr_block_storage refresh due to a regression introduced in v2.30.0 and persisting through v2.31.0.
✅ Acceptance Criteria
vultr/vultr provider constraint is = 2.29.0 in all four locations: envs/dev/main.tofu, modules/vultr/block_storage/main.tofu, modules/vultr/firewall/main.tofu, modules/vultr/instance/main.tofu
.terraform.lock.hcl regenerated with 2.29.0 hashes
Renovate config updated with a packageRule that excludes vultr/vultr versions > 2.29.0 until the regression is resolved upstream
tofu fmt -check -recursive opentofu/ passes
pr-tofu-plan-develop CI is green (no Plugin did not respond crash)
Regression boundary:vultr_block_storage ReadResource panic (Plugin did not respond) was confirmed on v2.30.0, v2.30.1, and v2.31.0. v2.29.0 is the highest clean version.
Module state: The null_resource attach workaround was removed in GHO-109 and does not come back with this change; we continue to rely on native attached_to_instance support.
Known side effect (separate follow-up): Upgrading from v2.28.1 → v2.29.0 causes vultr_firewall_rule resources to plan -/+ destroy and then create replacement because the provider changed how the source attribute is represented in state. Our module already uses subnet/subnet_size so configuration is correct; this is a one-time state-migration artifact. Will be handled in a follow-up (accept recreation with brief origin-firewall gap, or state rm+import to avoid).
Story Summary
As a platform engineer, I want the Vultr OpenTofu provider pinned to v2.29.0, so that plan CI doesn't crash on
vultr_block_storagerefresh due to a regression introduced in v2.30.0 and persisting through v2.31.0.✅ Acceptance Criteria
vultr/vultrprovider constraint is= 2.29.0in all four locations:envs/dev/main.tofu,modules/vultr/block_storage/main.tofu,modules/vultr/firewall/main.tofu,modules/vultr/instance/main.tofu.terraform.lock.hclregenerated with 2.29.0 hashespackageRulethat excludesvultr/vultrversions> 2.29.0until the regression is resolved upstreamtofu fmt -check -recursive opentofu/passespr-tofu-plan-developCI is green (noPlugin did not respondcrash)📝 Additional Context
vultr_block_storageReadResource panic (Plugin did not respond) was confirmed on v2.30.0, v2.30.1, and v2.31.0. v2.29.0 is the highest clean version.= 2.28.1(pinned in GHO-109 / PR feat(GHO-109): upgrade Vultr provider to v2.28.1, remove block storage attachment workaround #318)= 2.29.0null_resourceattach workaround was removed in GHO-109 and does not come back with this change; we continue to rely on nativeattached_to_instancesupport.Known side effect (separate follow-up): Upgrading from v2.28.1 → v2.29.0 causes
vultr_firewall_ruleresources to plan-/+ destroy and then create replacementbecause the provider changed how thesourceattribute is represented in state. Our module already usessubnet/subnet_sizeso configuration is correct; this is a one-time state-migration artifact. Will be handled in a follow-up (accept recreation with brief origin-firewall gap, orstate rm+importto avoid).📦 Definition of Ready
✅ Definition of Done