A Pulsar Cluster Where TTL and Retention Are Told Apart
A TDMQ for Pulsar cluster and its namespaces. msg_ttl is how long an UNACKNOWLEDGED message lives, not how long consumed ones are kept: too short silently drops work when a consumer is slow, too long turns a stuck consumer into unbounded backlog, and both extremes are refused. Retention is the separate thing that lets a new subscription read history.
Verification
Static-verifiedPassed: validated and lint-clean (provider-schema-validated for AWS/Azure/GCP; Terraform-language lint elsewhere).
Conformance
- Static validation (fmt · validate · tflint)
- No applicable security policies for this provider
- Plan tests (mocked: validation rules · outputs)
Provenance
- SHA-256 checksum
- Signature (pending)
Functional
- Live test pending (no cloud run yet)
Last verified 2026-09-15 · how we verify
Use it from the registry
terraform · opentofumodule "tdmq_queue" {
source = "www.iac-bazaar.com/iac-bazaar/tencent-tdmq-queue/tencentcloud"
version = "1.0.0"
}Needs a registry token from /account/tokens. The module itself is free; the account is what identifies you. Full setup: registry docs.
Inputs & outputs
Create a free account to read this module's contract
The declared contract - every input name, type, default and description, plus every output - is shown to signed-in accounts, not to anonymous visitors.
A free account sees the contract of every module in the catalogue. There is no subscription and nothing to buy - the modules are free to download, and they run under Vizier.
Documentation
tencent-tdmq-queue
A message broker on Tencent Cloud TDMQ for Pulsar. Works with Terraform and OpenTofu
(>= 1.6), tencentcloud provider >= 1.81.0, < 2.0.
msg_ttl is how long an UNACKNOWLEDGED message lives, not how long consumed messages are kept. Too short silently drops work when a consumer is slow; too long turns a stuck consumer into unbounded backlog. Both extremes are refused.
Retention and TTL are different things with similar names. Without a retention policy a subscription created tomorrow starts from tomorrow - correct, and not what people mean when they say they keep seven days.
A namespace is the unit of isolation and of quota, so an unrelated high-volume producer sharing one with your order pipeline is how one team's incident becomes two.
Verification
Static validation runs tofu fmt, init, validate, tflint and checkov.
This module has not yet had a live test, so it is published as statically
validated with its live test pending and does not carry the live-tested mark.
Usage code & full reference need an account
The complete copy-paste usage, the full input/output reference, and operational notes are free with an account - shown here and bundled in the download. Sign in and this section fills in.
- Usage
Related modules
tencent-ses
A Tencent Cloud SES domain, its addresses and its templates. The module exposes the exact DNS records Tencent is waiting for, because that is the thing you need next and it lives outside this Terraform. DKIM is on, since unsigned mail is accepted by the API, filed as spam and reported as sent; and a template is reviewed manually before it works.
aws-mq
deployment_mode defaults to SINGLE_INSTANCE, and Amazon MQ reboots the instance to patch it - so a single broker has planned downtime on AWS's schedule. Active/standby across two AZs, both log streams on, and passwords in a separate variable from users, because a sensitive value cannot be a for_each argument at all.
aws-kinesis-firehose
Without error_output_prefix, records Firehose could not process are written into the same prefix as the ones it could, wrapped in an error envelope that whatever reads the prefix treats as data. Nothing reports it.
huawei-kafka-streaming
A DMS Kafka instance with enable_auto_topic off, because a producer that misspells a topic otherwise creates one: the messages go somewhere real, nothing errors, and nobody consumes them. TLS and SASL are both on, which is what makes the user mean anything; the disk is encrypted at creation; and what happens when the disk fills is a decision you take.
oci-queue
dead_letter_queue_delivery_count defaults to zero, which is no dead-letter queue: a message a consumer cannot process is redelivered after every visibility timeout until retention expires, a poison message that holds a consumer for a day. Five deliveries then the dead-letter queue, seven days of retention instead of one, and a vault key instead of an Oracle-managed one.
oci-streaming
A public stream pool is an FQDN reachable from anywhere with a valid token; auto_create_topics lets any producer create a stream by writing to a new name; retention defaults to 24 hours, so a consumer a day behind loses data with no error on the producer side. Private endpoint behind NSGs, declared streams, seven days of retention, and the stream that loses data soonest reported as an output.