The performance is horrific over PPTP. I try using multiple using nth routing to distribute traffic (1 pptp, 2 uplinks, from my office to a data center, but even with turning off all those rules and just using one PPTP link it sucks.
Equiment is a Mikrotik 450 G on my end, cable modems (2) and a Mikrotik 1100 AH in the data center.
* The line supposedly is ok, everyone keps telling me.
* Speedtest.NET not using PPTP works fine, too. I get excctable bandwidth numbers as it should be.
* THe moment I put up PPTP things deteriorate.
The idea is to run all traffic over the PPTP to give better bandwidth, but at the moment it is in joke territory.
I am lucky enough to have a high quality bandwidth test application
I get a financial data feed from a company in the US and the client tracks everything, from jitter to packet loss.* Using non-pptp they start their historical downloads when I ask for an update of the local files with 7mbit, their max speed before their limit kicks in, the connection is very nice, no packet loss (0 per 1000 packets)
* Running over PPTP... I am lucky to get 100kbit download speed for the archive. I get network warning signs from their client. I get a packet loss measured of 1.1% - 2.5% which kills their transfer speeds, constant resend log entries.
* Running locally from the data center (same AH involved) - same as with no PPTP, perfect line.
Again, this is without nth and all that playing around
Just a PPTP connection between two mikrotiks.Anyone an idea? Traffic shaping for GRW from my host? I am at a loss how to get a better diagnose as I dont really have access to the equipment between (i.e. the internet).