Tahoe has problems, the worst of which are 32-on-64 problems, but it's also not mirrored, so I'm trying to move everything off of it. Tonight I moved the main webserver and the movable type server (which is called wiki.xen.prgmr.com, for historical reasons. our mediawiki server is book.xen.prgmr.com) I re-ip'd both, and both are running on both old and new servers until DNS finishes updating. I moved http to a brand new 64-bit image on boar at he.net- it should be much more reliable. I moved wiki to Coloma, a native i386-PAE box hosted at rippleweb in Sacramento.
Customers on tahoe are encouraged to move to new servers;
the samsung 750G drives (with 32M cache) are almost twice as fast as the segate 750G drives (with 16M cache) - even after removing the limiters (the seagates came with jumpers
that limited them to 1.5G sata1)
Anyhow, I've started to put system domains on the new server, Boar, and they are looking pretty good.
so we got the new servers installed and setup... and then we realised that disk was about 1/2 the speed we requested, and more importantly, there was serious problems with random access... copying /dev/zero to the disk would lock things up to the point where you couldn't even log in on another vt.
Obviously we're not putting customers on it until we figure it out; the server is in the garage right now for testing.
The server is going through burn-in as we speak
as I mention on the main page, we ran out of space the other day. we are putting in a new server, boar, and one of my ancient catalyst switches, with 'port monitor' or SPAN capabilities,
so I will be un-breaking the bridge on lion, and bandwidthD and my inward-facing IDS will both continue to function.
This will require us to physically re-configure the network (just moving cables- if we don't
screw it up, downtime should be less than 60 seconds- no reboot or anything, just a few dropped packets.)
Started whacking domains with the b&width hammer, as per the directions noted a couple entries ago.
Further refinements will probably involve putting quotas directly in
config files, with scripts to parse and automatically set limits at
domain creation. Ideally I'd also write a tool to re-assign domains to
existing classes.
This title is, of course, a complete lie. It looks like our disk layout scheme broke LVM snapshots. To quote from our testing:
# lvcreate -s -L 100M -d hydra_domU/test -n test_snap
Snapshots and mirrors may not yet be mixed.
That's some real well-supported technology there. Google gives me two results for that error message, both of which are source diffs.
I'm not sure what to do about this. Every so often I feel like abandoning LVM mirroring entirely and moving to LVM on MD, but that didn't exactly fill us with joy either.
I'm also considering bypassing the LVM-specific snapshot implementation and using the device mapper directly, but that worries me. I would want to know why snapshots and mirrors can't be mixed before implementing snapshots anyway.
Today I put my money in my mouth and worked on traffic shaping. I'm not 100% sure that this setup is correct -- we'll have to test it more before we put it in production. Tentatively, though, here's how it works:
We're doing everything in the dom0. Traffic shaping is, after all, a coercive technical solution. Doing it in customer domUs would be silly.
First, we have to make sure that the packets on xenbr0 traverse iptables:
# echo 1 > /proc/sys/net/bridge/bridge-nf-call-iptables
This is so that we can mark packets according to which domU emitted them. (There are other reasons, but that's the important one in terms of our traffic-shaping setup.)
Next, we limit incoming traffic. This is the easy part. To limit vif "baldr" to 1mbit /s, with bursts up to 2mbit and max allowable latency 50ms:
# tc qdisc add dev baldr root tbf rate 1mbit latency 50ms peakrate 2mbit maxburst 40MB
This adds a queuing discipline, or qdisc, to the device "baldr". Then we specify where to add it ("root",) and what sort of qdisc it is ("tbf"). Finally we specify the rate, latency, burst rate, and amount that can go at burst rate.
Next we work on limiting outgoing traffic. The policing filters might work, but they handle the problem by dropping packets, which is. . . bad. Instead we're going to apply traffic shaping to the outgoing physical Ethernet device, peth0.
First, for each domU, we add a rule to mark packets from that network interface:
# iptables -t mangle -A FORWARD -m physdev --physdev-in baldr -j MARK --set-mark 5
Here the number 5 is an arbitrary integer. Eventually we'll probably want to use the domain id, or something fancy. We could also simply use tc filters directly that match on source IP address, but it feels more elegant to have everything keyed to the domains "physical" network device. Note that we're using physdev-in -- traffic that goes out from the domU comes in to the dom0.
Next we create an HTB qdisc. We're using HTB because it does what we want and has good documentation (available at http://luxik.cdi.cz/~devik/qos/htb/manual/userg.htm .) We won't go over the HTB options in detail, since we're just lifting examples from the tutorial at this point:
# tc qdisc add dev peth0 root handle 1: htb default 12
Then we make some classes to put traffic into. Each class will get traffic from one domU. (As the HTB docs explain, we're also making a parent class so that they can share surplus bandwidth.)
# tc class add dev peth0 parent 1: classid 1:1 htb rate 100mbit
# tc class add dev peth0 parent 1: classid 1:2 htb rate 1mbit
Now that we have a class for our domU's traffic, we need a filter that'll assign packets to it.
# tc filter add dev peth0 protocol ip parent 1:0 prio 1 handle 5 fw flowid 1:2
At this point traffic to and from the target domU is essentially shaped. To prove it, we copied a 100MB file out, followed by another in. Outgoing transfer speed was 203.5KB/s, while incoming was about 115KB/s.
This incoming speed is as expected, but the outgoing rate is a bit high. Still, though, it's about what we're looking for. Tomorrow we'll test this with more machines and heavier loads.
so, having rdns point back to me makes handling abuse reports much easier (that is, it makes it much more likely I will get the complaint rather than my upstream) - so I am going to require you to stay on a .xen.prgmr.com rdns until you have been a paying customer for 3 months.
Like everything, exceptions can be made, but if I don't know you, it's three months (or you can pay up-front for three months, with the understanding that you won't get it back if I shut you down for AUP violations.)
the prgmr.com AUP:
http://prgmr.com/aup.html
pretty standard, except for the bit where I prohibit all bulk mail without my approval.
I'm not interested in hosting even most double-opt in lists- most of the larger lists, even if they are legitimately double-opt in, generate more complaints than I am willing to deal with at these prices. If you are a legitimate mail sender, I would suggest you start with http://isipp.com
Recent Comments