| TX OVR errors on venet [message #50587] |
Tue, 17 September 2013 08:57 |
_sgb_
Messages: 3
Registered: February 2009
|
Junior Member |
|
|
Hi all,
I'm seeing what looks like intermittent dropped packets on venet0 on the hardware node. What could cause this? Packet TX problems on a virtual interface suggest a resource issue to me, but I'm not seeing any increase in the failcnt on beancounters.
The packets are destined for another container on the same machine. There is no netfilter setup running on any container or on the host node. We haven't introduced any bespoke routing rules.
In the interest of full disclosure, this is a proxmox ve setup and I haven't seen anything like this on my pure OpenVZ nodes, however this machine is more heavily loaded than any other node on our LAN. It's running approximately 25 RHEL5.5 containers.
The issue was noticed when we started to experience SNMP issues - intermittent 'ICMP - Destination Unreachable (Port Unreachable)' in the SNMP conversation, specifically after the 'get-response' packets. I'm not convinced that this is related because the packet drops don't match the ICMP timings.
When the issue was raised with IT, we took the majority of the other containers out of service (they weren't serving anything crucial at the time) in order to reduce the load on the networking stack to see if it made a difference. It doesn't seem to have. We're still experiencing increases in the TX OVR count:
root@vzdemo01:~# netstat -i
Kernel Interface table
Iface MTU Met RX-OK RX-ERR RX-DRP RX-OVR TX-OK TX-ERR TX-DRP TX-OVR Flg
eth0 1500 0 38164657 0 0 0 20615000 0 0 0 BMRU
lo 16436 0 3998296 0 0 0 3998296 0 0 0 LRU
venet0 1500 0 127397460 0 0 0 127380657 0 484 0 BOPRU
vmbr0 1500 0 37144791 0 0 0 19698118 0 0 0 BMRU
root@vzdemo01:~# vzctl --version
vzctl version 4.0-4.git.162dded
root@vzdemo01:~# uname -a
Linux vzdemo01 2.6.32-16-pve #1 SMP Mon Oct 22 08:38:13 CEST 2012 x86_64 GNU/Linux
root@vzdemo01:~# free -m
total used free shared buffers cached
Mem: 7936 7856 79 0 437 5515
-/+ buffers/cache: 1903 6032
Swap: 4095 357 3738
root@vzdemo01:~# netstat -an | wc -l
258
Any recommendations on troubleshooting steps would be very welcome. I'm considering offloading some checksumming to the physical card but I'd prefer to figure out what's happening before then.
Best regards,
Simon
[Updated on: Tue, 17 September 2013 09:11] Report message to a moderator |
|
|
|
OpenVZ Forum
Members
Search
Help
Register
Login
Home
