Hi @EricPerl 

Thanks for posting in our business forum.

EricPerl wrote

  @Clive_A ,

 

By the way, I suspect none of these limits are documented (designed for 60-80, hard fail at ~250).

I also suspect the failure didn't generate any logs. Wouldn't it be nice if end customers could diagnose some of these failures?

Typical values for regular models as we have tested with the capacity in our trial manufacturer phase. And this is not certain as your environment might be different. How would it be documented as you are already being told about the session's upper limit? That'd be something you should pay extra attention to.

TBH, it's very picky to say that it is not documented. There is no need to be that specific about exactly how many clients are reached.

One could use up all the 15,000 sessions on a single server in extreme circumstances in a business environment. Will that be misleading if you document it as recommended for 60 devices or so? Extra further explanation would be needed.

Picking up the devices for your network requires some basic network knowledge or if you don't wanna pay to learn, then browse as many forums and tech channels for extra tips. Products from different vendors can really vary in different user cases. Like we used to only provide ER605 and 7206. 8411 was the third one and later projected for the higher needs of our customers. And your network scale or level is expanding gradually unless you pay someone to fix high-end gears for you to be future-proof when you provide enough details about your future plans and growth.

I mean you have to know how things work and most vendors do before you make a judgment on A B and C.

If you carry 100 IoT devices that do not heavily exploit your network for another mining thing (if there is a brand), it should be okay to use them. IoT is not network heavy, 60-80 is considered as regular devices like phones and PCs.

I think this has been explained at least 5 times in detail on the forum since I am responsible for the routers.

If the DHCP got a problem, you have the option to display DHCP-related logs for troubleshooting. For pro users, it is not hard to build a syslog with more details. Syslog has long existed in networking and should be the one that plays an important role in a network.

I don't deny you can integrate everything into one product. That will inevitably speed up the cycle of a product(version). OC200 V2 has been discussed if you know some news about it. The performance is throttled and we have to remove some performance-heavy features to keep it working. (Also explained this so many times about the product design. I know someone would argue that we should provide more powerful devices at cheaper prices. That's basically impossible. The cost of better hardware will inevitably increase the tag price. This is how everything works.)