Open dhiaayachi opened 3 weeks ago
Thank you for reaching out! I understand you're looking to minimize the impact of API calls on the namespace replication queue bandwidth.
It's a great idea to prioritize bandwidth for failover messages. Here's a suggestion:
You can leverage the namespace.replication.max-message-per-second
setting to limit the rate at which messages are replicated.
This helps ensure that failover messages get the bandwidth they need. You can find more details about this configuration setting in the Temporal server configuration.
Let me know if you have any other questions.
Thank you for reaching out and bringing this to our attention! This is an important consideration for efficient and reliable cluster failover. While we don't currently have a built-in mechanism to specifically throttle API calls for the namespace replication queue, there are a few approaches you can consider:
golang.org/x/time/rate
. This allows you to control the number of requests sent per time period. Please let me know if you have any further questions or need assistance with these approaches.
Thank you for reaching out! I understand that you're concerned about potential API spamming and want to keep the namespace replication queue's bandwidth low for cluster failover messages.
While there isn't a dedicated feature to explicitly control the rate of API calls for namespace replication, here are a few suggestions to address your concerns:
replication_task_poll_interval
to control the rate of namespace replication tasks. Adjust this setting to suit your bandwidth requirements. replication_filters
configuration to selectively filter out namespace replication messages for specific types of workflows or tasks, reducing the overall replication traffic. I hope these suggestions help manage your namespace replication traffic. If you have any further questions or need additional assistance, please feel free to ask.
Need to ensure the API isn't spammed since we want the namespace replication queue to be relatively low bandwidth for cluster failover messages.