| Awfully slow network performance w/ r8169 and openvz [message #41105] |
Fri, 19 November 2010 07:17  |
felicitus
Messages: 1
Registered: November 2010
|
Junior Member |
|
|
Hi,
I'm currently setting up my first openvz server. It seems that my network speed is limited with the openvz kernel and r8169 (RTL8111/8168B PCI Express Gigabit Ethernet controller). I tried both 2.6.27-levitan.1 and 2.6.32-dzhanibekov.1.
I did not set up any containers so far; all things mentioned in here refer to VE0!
If I boot my server in a rescue system (stock debian system with stock kernel), I receive maximum speed (more than 10mbyte/s for a remote download). However with openvz, my performace is stuck at 400-600kbyte/s, and drops down to below 100kbyte/s for a download which takes a bit longer (let's say 100MB).
I have:
* Gentoo (current version), vzctl 3.0.23-r2
* No messages in the kernel log with errors or network issues
* No iptables set
* No special configuration like bandwidth limiting or such
Negotiation is OK:
arkus conf # mii-tool
eth0: no autonegotiation, 100baseTx-HD, link ok
No packet errors:
eth0 Link encap:Ethernet HWaddr 6c:62:6d:76:2f:2d
inet addr:hidden Bcast:hidden Mask:255.255.255.192
inet6 addr: hidden Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:111925 errors:0 dropped:0 overruns:0 frame:0
TX packets:90386 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:158226334 (150.8 MiB) TX bytes:6913452 (6.5 MiB)
Interrupt:249 Base address:0xc000
/proc/user_beancounters:
arkus conf # cat /proc/user_beancounters
Version: 2.5
uid resource held maxheld barrier limit failcnt
0: kmemsize 7284234 13681307 9223372036854775807 9223372036854775807 0
lockedpages 0 3075 9223372036854775807 9223372036854775807 0
privvmpages 21592 92826 9223372036854775807 9223372036854775807 0
shmpages 1310 1326 9223372036854775807 9223372036854775807 0
dummy 0 0 9223372036854775807 9223372036854775807 0
numproc 175 246 9223372036854775807 9223372036854775807 0
physpages 4290 61920 9223372036854775807 9223372036854775807 0
vmguarpages 0 0 9223372036854775807 9223372036854775807 0
oomguarpages 4290 61920 9223372036854775807 9223372036854775807 0
numtcpsock 16 19 9223372036854775807 9223372036854775807 0
numflock 2 18 9223372036854775807 9223372036854775807 0
numpty 2 3 9223372036854775807 9223372036854775807 0
numsiginfo 2 56 9223372036854775807 9223372036854775807 0
tcpsndbuf 628864 628864 9223372036854775807 9223372036854775807 0
tcprcvbuf 262144 361368 9223372036854775807 9223372036854775807 0
othersockbuf 30056 117152 9223372036854775807 9223372036854775807 0
dgramrcvbuf 0 5648 9223372036854775807 9223372036854775807 0
numothersock 57 78 9223372036854775807 9223372036854775807 0
dcachesize 771540 872256 9223372036854775807 9223372036854775807 0
numfile 3003 3998 9223372036854775807 9223372036854775807 0
dummy 0 0 9223372036854775807 9223372036854775807 0
dummy 0 0 9223372036854775807 9223372036854775807 0
dummy 0 0 9223372036854775807 9223372036854775807 0
numiptent 0 0 9223372036854775807 9223372036854775807 0
Output of tc qdisc:
arkus ~ # tc qdisc
qdisc pfifo_fast 0: dev eth0 root bands 3 priomap 1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1
I'm not too sure if the issue is maybe related to bean counters; but as the fail count is 0, I assume everything is correct here?
arkus conf # uname -a
Linux arkus 2.6.27-openvz-levitan.1 #1 SMP Fri Nov 19 07:12:32 CET 2010 x86_64 Intel(R) Core(TM) i7 CPU 930 @ 2.80GHz GenuineIntel GNU/Linux
arkus ~ # tc class show dev eth0
<no output returned>
Anything I could check else?
Thanks
Felicitus
|
|
|
|
|
|
|
|
|
|
OpenVZ Forum
Members
Search
Help
Register
Login
Home
