Hey everyone, We have georedundant servers in production, and due to the delay between them, we are encountering some errors with a database. Now, we are unable to replicate in our QA lab because everything is located in the same building. To resolve this, we want to use tc. Basically what I need to do is use tc to add a 15 ms delay on egress for anything within the destination subnet of the "georedundant" site. Likewise, we want to add an ingress delay for anything from the georedundant site. Now, I was planning to use HTB for this, however when I follow the manpage for htb it says that I *must* specify the rate for each of the child classes. The thing is, when I try to specify a rate of 1000mbps (the line speed) I get an error "HTB: quantum of class 10001 is big." I am planning on having a root class, and two child classes, one for the georedundant subnet, and one for the default. However if the georedundant is being used, I want it to use the full bandwidth available on the line, I don't want to limit it - we simply have too much data going across (2x1Gbps links running ALB). So can anyone advise on how to get HTB to accept that rate? The only documentation I've found thus far refers to calculating burst, and a lot of their numbers aren't explained very well. The documentation isn't very helpful. -Matthew -- To unsubscribe from this list: send the line "unsubscribe lartc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
