storyboardassociates wrote:
I hope the next revision of the firmware resolves all these issues, otherwise DrayTek does not deserve to claim any credentials for enterprise grade solutions

It was 2 years ago I started this thread about firmware issues on the 3910 and in that whole time we have only seen a couple of stable builds. When you consider this was around £800 when first released that is just not good enough.

FYI I use Untangle in other company installations with this. Fantastic device.
https://www.supermicro.com/en/products/system/Mini-ITX/SYS-E300-9D-4CN8TP.cfm

But to tell the truth, the 3910 (no IPv6 configured, 100/100 microwave link, 3 VLANS, VPN) runs just smoothly and rock solid in a really large company (at least 120 employees with almost two devices each)

https://postimg.cc/MMc9wJ5L

qwaz01 wrote:

storyboardassociates wrote:
I hope the next revision of the firmware resolves all these issues, otherwise DrayTek does not deserve to claim any credentials for enterprise grade solutions

It was 2 years ago I started this thread about firmware issues on the 3910 and in that whole time we have only seen a couple of stable builds. When you consider this was around £800 when first released that is just not good enough.

Can't remember how long it took Draytek to maker the 2960 stable. However two years is far too long although it was a cheap high end product. Mine seems stable at the moment although I got to admit any changes I make don't need a reboot I reboot anyway.

I see 4.3.2_RC5a is now on the international site for the 3910. Has anyone been brave enough to try it to see if the memory leak is still present?

Not yet but I've been running 4.3.2_RC8 on the 2962 because all the bug fixes listed in that firmware are also present on the 3910 - System uptime of 434:29:00, Memory usage sat at 69% but I'm only using IPSEC site-to-site, no SSL-VPN or Wireguard (which appears to be where the memory leak lies judging by past comments)

The following are both issues I'm having with a 3910 at multiple sites. The untagged devices getting IP's from networks with VLAN tags has been driving me bloody bonkers, I thought it was a switch configuration issue, I never suspected the 3910 - I was furious when I read the below in the release notes (i've complained about the load balance policy issue in this thread too)

Corrected: An issue with intermittent packet loss when routing through load balance policy using IP alias on a HA setup
Corrected: An issue where an untagged PC got an IP from a network that had a VLAN tag.

Been running 4.3.2_RC5a for a day so far without issue but I am not using SSL-VPN or Wireguard.
Memory usage started at 64% and settled to about 71% over the course of 12 hours so far. It had been sat at 74% memory usage on the Dashboard with 60+ days uptime previously so not particularly concerned.
No issues so far but I'll report back if anything crops up.