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

[jkim86]

Anybody else seeing a pattern with 50E dying? I had these deployed for about two years. I've had two starting to have packet loss and one that's not forwarding even though port is up. I saw a thread on reddit and this is starting to become a growing concern. I was able to have the client replace two of them but I'm afraid we're going to start seeing more problems with the 50E. Anybody know if we can RMA these still without support? The client is going to lose confidence in Fortinet. This rate of failure is unacceptable.

Update 2022-02-03:

Had another two breaking down. 1 is showing slow speeds. The other one isn't passing traffic even though port is up. 6th 50E broken now.

Update 2022-04-26:

2 more units bad.

[AlexC-FTNT]

Hardware support is needed to be able to open a case that leads to RMA, so no, you can't RMA the unit without support. First we need to prove that packet loss - so check first that the unit is not enetring conserve mode, or that a hardware test (HQIP test) has been performed to prove the RMA. Simply stating that "we see packet loss" without actual proof is not enough for RMA.

[jkim86]

ya just starting to see a trend here since 2 units suddenly have packet loss when they weren't having any problems before and a 3rd unit that stopped working on lan1. All reported in the last two months. Also reading other users on reddit experiencing similar problems with his 50E and already replaced 4 last year.

[jkim86]

I'll run the HQIP tests once I replace the units from the remote sites and get them back.

[yashwani]

HQIP tests are required to confirm issues with hardware, alternatively if TAC confirms for RMA after troubleshooting. 

[jkim86]

Had another two breaking down. 1 is showing slow speeds. The other one isn't passing traffic even though port is up. 6th 50E broken now.

[bezeqsoc]

Same problem.
Have to replace 5 units in the last 2 month during packet loss issue.

All ours are with different versions but above 6.x.x

Could it be a software problem?

[AlexC-FTNT]

Yes and no. The new firmware comes with new features, but the hardware specs on the 50E remain unchanged. So you may see that 6.2 version doesn't have enough space on flash memory to run (diag sys flash list) and that causes crashes, reboots, conserve modes, etc (ultimately seen as 'packet loss').

You can ask for replacement, but RMA is not the solution. You need to run FortiOS 6.0, or  a very minimum configuration in newer version (no GeoIP, extended IPS database, etc). FortiOS 6.2.11 will have a fix to that (can't give technical details behind it, but I assume some features will be disabled in order to make it run stable)

[bezeqsoc]

In my case I have packet loss with no uplink connection - just workstation connected directly to the unit after factory reset There is no packet loss when connected to wan1/wan2 interface
Looks like lan ASIC is dead or something...

Also, especially downgraded to  5.6.14 - no effect, the problem persist.  

[jkim86]

Interesting. I got some units back and I also found the WAN ports are working.