Please confirm if there is any restriction on the number of subscriptions topics per Non-durable Queue level for solace appliance 3530?
I believe it's 50,000. Shouldn't be too hard to test..!
Thank you @Aaron
Kindly confirm below details for Solace PubSub+ 3530
a. Please confirm if there is any restriction on the topic subscription count per Non-durable Queue?
b. Is their the total number of subscription limitation at box or VPN level, but just wanted to check if there is one or not for per Temp Queue level. The one highlighted in orange.
c. What if the topic subscription count limit is 50000 and we increase topic subscription count to 50000 for 1 queue then will it breach the limit or stop other client to connect?
If yes, then what preventive action can be taken.
d. 1. Number of topic subscriptions that can be assigned to a queue?
2. Number of topic subscriptions that can be assigned to all queues?
Ok, so I checked the limits and alerts spreadsheet... for a 3530, it is 100k max for the whole appliance. And it is also up to 100k per queue. So you can add 100k onto one queue, or 50k onto two queues, or however else you want to do it. Doesn't matter durable or non-durable (temporary) queues.
This limit goes to 500k if you have 3260 or 3560 with a larger NAB: 6x10GbE or 8x10GbE.
I'm assuming if you have appliances, then you're a Solace customer with paid support. You can view the limits and alerts spreadsheet yourself via https://products.solace.com Look under "Appliance" --> version --> etc. E.g.: https://products.solace.com/products/10.3-Production/Appliance/Current/10.3.1.17/
I don't see anything highlighted in orange, so not sure what you're referring to?
Any client or admin command trying to add the 100,001st topic subscription to a queue should get an error from the broker, but will not affect any existing subscriptions.
Total number of subscription on endpoints can be monitored using `show message-spool detail` in CLI:
Or its equivalent in SEMPv1: <rpc><show><message-spool><detail/></message-spool></show></rpc>
Hope that helps.