I am attaching a screenshot once again of an ongoing problem that I am having. I have several other routers setup in the same fashion as this one, but this one in particular is giving me fits. If you look in my screen shot I have a ton of TX drops on Vlan interfaces.
The only thing I can think of that is any different on this machine than any of my other ones is that there is a Cisco Layer 3 switch in between my router and my offsite switches over the Fiber. I do not have access to that switch as it belongs to our ISP for the fiber. I have been troubleshooting with the Admin from the ISP however nothing appears to be wrong. We do not use Auto Negotiate, we are set to 1Gbps Full.
I have the e1000 driver set in vmware now, it was set to flexible and it was causing the machine to crash @ 100% cpu usage. I was thinking maybe it is a bad network card or just this version of RouterOS we are using.
Any thoughts, maybe MTU…not sure that would cause the TX drops.
I have several other Routers, same setup in general, the difference is ROS versions.
Examples of my x86 machine versions:
4.11 - Has the same number of Vlans as the v5.1 in question, no problems.
4.30 - Not as many Vlans but using hotpots.
4.60 - Not as many Vlans but more traffic and pps.
I have been considering actually going backwards in versions and going back down to like v4.11 or something.
Also the hardware this is on is an i2600, 8GB Ram. Others have reported some issues with v5 of ROS, one of which I actually replied to and one of which I started and thought I had fixed.
my high cpu was being caused by routerOS installed as a virtual machine in ESXi 4.1 with the adapter type as flexible. I fixed the high cpu load by changing it to e1000. The high TX drops I was never 100% convinced were a huge problem, but I do queue my users as a total using the queue tree, capping the total out @ 200Mbps and limiting each user on a /24 to 6Mbps/2Mbps. Some of the dropped TX packets were looking to be caused by queues.
We use queues too. But not queue tree, only simple queues.
Due to TX drops on vlans we get an ICMP package loss and traffic failing every 1-2 minutes from 35 to 8 mbits
Did anyone ever find a specific fix for the tx-drops only on vlans? running current v6 and we’re experiencing the issues using decent hardware with intel e1000 drivers