New Features: Queue Alerts and Inbound Queue Management
Two new capabilities land in this release. Both give you more control over message queues - one for spotting them, one for clearing them.
1. Queue Alerts for Production and QA
You can now configure alerts when an inbound or outbound queue builds up - in both Production and QA environments.
Alongside the existing Warnings and Errors notifications, each queue type now has two settings:
- Threshold (minutes) - how long a queue must persist before the first alert fires
- Re-alert interval (minutes) - how often to remind you while the queue stays active
Configure inbound and outbound independently. Configure Production and QA independently. Different thresholds for different risk levels, no problem.
A note on suspended endpoints: Queue alerts cover all endpoints in your subscription, including suspended ones. This is deliberate. An endpoint suspended by accident will still receive inbound messages, and the queue will start building silently. The alert makes sure you notice.
If you're seeing queues build up on a suspended endpoint, that's a signal: either the endpoint shouldn't be suspended, or the sending system needs to stop routing messages to it. The cleanest fix is to delete the endpoint entirely so messages have nowhere to land.
Why it matters: Inbound queues form when messages arrive faster than they can be processed - traffic spikes, batch imports, downstream slowdowns. Outbound queues form when a target system stops responding cleanly. Both used to sit quietly until something downstream broke. Now you see them early.
How to set up alert notifications about integration issues for both QA and Production environments
2. Inbound Queue Management
When an inbound queue needs intervention, you can now handle it yourself without raising a support ticket.
From the Endpoint Status view, you can:
- Open queue details directly
- See queue count, oldest message age, and a summary grouped by entity
- Inspect the content of the oldest queued message
- Suspend the endpoint, then choose how to clear the queue:
- Remove only the oldest message
- Remove all messages of a specific entity type
- Remove all messages in the queue
Why it matters: Sometimes a stuck message blocks everything behind it. Sometimes a single bad entity type is poisoning the queue. Sometimes the whole batch needs to be flushed. You shouldn't need to file a ticket and wait. You should be able to look, decide, and act.
Identifying and Managing an Inbound Queue
We'd love your feedback! If you have questions or ideas for further improvements, let us know in the comments below.
Comments
Hi Janne,
thank you for sharing this, glad to see those improvements!
I think the second link (https://support.oneio.cloud/hc/en-us/articles/27579506133020-Identifying-and-Managing-an-Inbound-Queue) doesn't work, at least for me.
BR,
Sebastian
Hi Sebastian, thank you for your update! Can you see it now?
Hi Tatiane,
yes, it works, thank you.
Please sign in to leave a comment.