Engenius ECS2512FP $300 (offer) - 8x2.5GbE POE++ and 4xSFP+

[T_Minus]

I could have an extra one to sell you for cost. If needed I can plug an empty one into a killawatt and tell you what it reads.

I may take you up onthat! I'll send you a pM

Their webpage calls it "cloud managed" but I don't want the cloud for anything with my home (or any) network... can I manage it 100% local with no cloud anything?


@bvd thanks for power info and screenshots!!

 

[cybrnook]

I may take you up onthat! I'll send you a pM

Their webpage calls it "cloud managed" but I don't want the cloud for anything with my home (or any) network... can I manage it 100% local with no cloud anything?


@bvd thanks for power info and screenshots!!

Sounds like a plan, I'll keep an eye out for it.

And yes, as long as you have the ability to stand up a Ubuntu VM you can use the "controller" to manage it for free (Software Packages (VM) - FitController User Manual). If not, it still has the standard web management page behind it's IP on the network, you can always log into with un/pw and manage it directly without any controller (or cloud connect).

 

[bvd]

I may take you up onthat! I'll send you a pM

Their webpage calls it "cloud managed" but I don't want the cloud for anything with my home (or any) network... can I manage it 100% local with no cloud anything?


@bvd thanks for power info and screenshots!!

No problem! And the gif I attached, that's all from the local UI

I mentioned in the original post, they can be managed by cloud, local UI, or both (hybrid) - a lot of the functionality is actually only available via the local UI, including virtually *all* of the more advanced features like port forwarding/blocking, policy based routing (generous use of the word 'policy' there lolol).

You don't 'need' anything to manage it - the fitcontroller VM is just so you can manage APs and switches (and soon firewalls as well, once they release em) in the same place, but it's still subject to the same feature limitations as their cloud solution.

 

[bvd]

Sounds like a plan, I'll keep an eye out for it.

And yes, as long as you have the ability to stand up a Ubuntu VM you can use the "controller" to manage it for free (Software Packages (VM) - FitController User Manual). If not, it still has the standard web management page behind it's IP on the network, you can always log into with un/pw and manage it directly without any controller (or cloud connect).

You dont need a VM btw - you can run it on anything that can run docker

 

[cybrnook]

You dont need a VM btw - you can run it on anything that can run docker

I figured as much….. was already scratching my head on how to cherry pick the install script to turn it into an Unraid template, but ended up not pursuing it.

 

[Dave Corder]

Does anyone happen to have an actual FIT access point (preferably connected to a local FitController, but not sure that's strictly necessary)?

I've been tinkering tonight with my newly acquired EWS850AP and try to convert it to an EWS850-FIT and register it in a local FitController I have running under Proxmox. I think I was able to convert it and get it flashed with the EWS850-FIT firmware (it now shows as EWS850-FIT in the web UI, however the overall skin of the web UI is the same as it was under EWS850-AP, so I'm not 100% positive).

However, I've hit a stumbling block in that the controller seems to reject my device's serial number when I try to register it. I'm wondering if maybe actual FIT devices have a certain serial number format that the controller validates against, so I hope to be able to get someone else's legit FIT serial number to compare to so I can figure out where to go from here.

 

[bvd]

Does anyone happen to have an actual FIT access point (preferably connected to a local FitController, but not sure that's strictly necessary)?

I've been tinkering tonight with my newly acquired EWS850AP and try to convert it to an EWS850-FIT and register it in a local FitController I have running under Proxmox. I think I was able to convert it and get it flashed with the EWS850-FIT firmware (it now shows as EWS850-FIT in the web UI, however the overall skin of the web UI is the same as it was under EWS850-AP, so I'm not 100% positive).

However, I've hit a stumbling block in that the controller seems to reject my device's serial number when I try to register it. I'm wondering if maybe actual FIT devices have a certain serial number format that the controller validates against, so I hope to be able to get someone else's legit FIT serial number to compare to so I can figure out where to go from here.

I haven't bought one, but as I recall it, the SN used was something you could only view via the local UI of the Fit device - it's a longer SN, similar to the ECW line, but wasn't on the sticker... I'll try to find my notes on it tomorrow night after family leaves town, I'm pretty sure I've got a SN jotted down too actually, but I'll let ya know

 

[Dave Corder]

I haven't bought one, but as I recall it, the SN used was something you could only view via the local UI of the Fit device - it's a longer SN, similar to the ECW line, but wasn't on the sticker... I'll try to find my notes on it tomorrow night after family leaves town, I'm pretty sure I've got a SN jotted down too actually, but I'll let ya know

Thanks. Any info you have would be helpful. I stayed up way too late last night mucking around trying to reverse engineer the Fit Controller serial number validation stuff. With a little sleuthing on Google, I was able to dig up two valid serial numbers for FIT devices: 2330GC11FMX9 for a EWS2910P-FIT and 2380XC31WH5N for a EWS850-FIT (edit: and 22C0X451F8NG for EWS377-FIT). My guess at the moment is that the first four characters are a model code prefix of some kind, the next 7 are unique, and the last character is a check character of some kind. I did find a refresh to a hash table with characters and their codes, but I don't know what the check character algorithm is yet.

 

[bvd]

My guess at the moment is that the first four characters are a model code prefix of some kind, the next 7 are unique, and the last character is a check character of some kind. I did find a refresh to a hash table with characters and their codes, but I don't know what the check character algorithm is yet.

I think the first two are likely the manufacturing year? Or does that not make sense based on the timeline for whenever the posts you found these on came up maybe?

My feeling is that, with enough "model + serial + mac address" combos, we likely should be able to crack this - they tend to use device addresses for a lot of things, at least historically speaking anyway.

I'm out in the boonies right now over at my in-laws (where internet connectivity is still measured in kbps ), but depending on the kiddos, hoping to be back home tonight before I've gotta pass out (work tomorrow and Saturday) and have a little time to do some digging!