Fortigate 50E Dying. Packet loss and lan ports stop forwarding. Update 2022-02-03 6th unit.

Now the next 50E has died. So far we've broken 5 out of 18 50Es with this bug. Devices from the batches with the serial numbers FGT50E3U16... and FGT50E3U17.... are affected. We are now initiating RMA processes for 4 devices. It would be nice if Fortinet would make a basic statement about an existing series problem.

I have 2 devices within Serial Range FGT50E561801.... A statemement from fortinet would be  welcome.

I have the same problem with a serial number: FTG50E3U16. At the beginning of the year, the ports became slow, then there were more and more failures, and since last month, nothing works at all. Even resetting the firewall did not help. Unfortunately, the warranty has already expired. This is very frustrating. I expect Fortigate to provide a statement on this issue, as I suspect it is a case of planned obsolescence.

hurray 2 more dying.

I have also one more dying last week (very low speed and packet loss). 3rd one in 3 Months. 

Just got two 51E having the same issues (WAN OK, all LAN port showing packet loss)

In this week, both are proceeding RMA.

I am afraid there will have a lot upcoming...

We have seen several of these appear over the past 6 months so I have opened one and had a look at it (I am an electronics engineer and like to fiddle more than I should!).

My dead 50E is now alive and kicking having reheated the main CPU which is a small BGA device under the heatsink. (For history this is a common problem with BGA devices which dont have enough cooling, especially following the introduction of lead free solder 10 years ago.  This results in slight expansion of the cpu / board and mild warping and then the previously flat cpu sitting on a bed of solder balls develops a dry joint and thus a bad connection).

I ran  ping test from the previously unresponsive LAN side and saw 400 packets with 1 dropped packet and a fully responsive web interface (this was completely dead in this instance previously). I have also updated the firmware and will soak test this now to see how it perfoms over a few days / week. 

@Rob_Roche Do you have before and after pics?

Not much difference to be seen, however here is a pic from after it was heated and was running - CPU highlighted. (yes I know it should have a heatsink on it - was testing in the immediate aftermath)

Hi jkim86,

The problem of insufficient flash space of 50E will cause the device to fail to work normally. The ISSUE should match this. It is recommended to upgrade to 6.2.14, and then release some flash space in the following release notes so that low-end models such as FGT30/50 can work an keep up the good work, thank you.

https://docs.fortinet.com/document/fortigate/6.2.14/fortios-release-notes/413989/fortigate-30e-and-50e-flash-card-space-optimization