Details
-
Type: Bug
-
Status: Closed
-
Priority: Major
-
Resolution: Fixed
-
Fix Version/s: OpenVZ-legacy
-
Component/s: Containers::Kernel
-
Security Level: Public
-
Environment:Operating System: Other
Platform: All
-
External issue URL:
-
External issue ID:2061
Description
Hello,
on the RHEL5 OpenVZ kernel you can run the netstat command and get info on connections for all processes and VPS's on the node. This is very useful in tracking down bandwidth hungry processes.
On the new RHEL6 kernel it seems the netstat command when run on the node no longer gives connections for all VE's on the node.
Can this please be changed so that when the netstat command is run on the node it will show all connections for all VE's.
Why its important:
1) using the netstat command on the node you can find all active connections based on IP, port and PID using the command:
netstat -n --program | grep ESTABLISHED
2) This information gives you PID and PORT relationships of all active network connections
3) Then using iptables with verbose output you can add new watches on these IPs and Ports and find out their specific bandwidth rate - which can then be assigned to the PID (application and VEID).
4) So in effect you can determine PID bandwidth usage in all VPS's running on the node - VERY IMPORTANT and USEFUL information to determine run away processes and which are the most bandwidth hogging applications.
Thanks.
on the RHEL5 OpenVZ kernel you can run the netstat command and get info on connections for all processes and VPS's on the node. This is very useful in tracking down bandwidth hungry processes.
On the new RHEL6 kernel it seems the netstat command when run on the node no longer gives connections for all VE's on the node.
Can this please be changed so that when the netstat command is run on the node it will show all connections for all VE's.
Why its important:
1) using the netstat command on the node you can find all active connections based on IP, port and PID using the command:
netstat -n --program | grep ESTABLISHED
2) This information gives you PID and PORT relationships of all active network connections
3) Then using iptables with verbose output you can add new watches on these IPs and Ports and find out their specific bandwidth rate - which can then be assigned to the PID (application and VEID).
4) So in effect you can determine PID bandwidth usage in all VPS's running on the node - VERY IMPORTANT and USEFUL information to determine run away processes and which are the most bandwidth hogging applications.
Thanks.
Attachments
Issue Links
- cloned to
-
OVZ-6600 [vz7] netstat command restricted
- Verified