[lvs-users] Not all subinterfaces come up after pulse restart

Will Liu (wiliu2) wiliu2 at cisco.com
Mon Jun 26 23:46:54 BST 2017


Hello,

I hope someone has seen this issue and can help.

Problem:
Each time I run "service pulse restart", ifconfig outputs only 1-3 of the 4 interfaces. So, randomly, not all the configured interfaces in the lvs.cf come up.

Here's what I mean to be more clear (ifconfig output, note em1:5, em1:6 may come up or a combination of up to 3 will show, even though I have 4 configured):
# ifconfig

em1:5     Link encap:Ethernet  HWaddr D4:BE:D9:EF:2F:D8
          inet addr:x.x.x.202  Bcast:x.x.x.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
em1:6     Link encap:Ethernet  HWaddr D4:BE:D9:EF:2F:D8
          inet addr:10.95.108.203  Bcast:10.95.108.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

Here is my lvs.cf:
serial_no = 300
primary = x.x.x.47
service = lvs
backup = 0.0.0.0
heartbeat = 0
heartbeat_port = 539
keepalive = 6
deadtime = 18
network = nat
nat_router = x.x.x.130 em2:1
nat_nmask = 255.255.255.224
debug_level = NONE
virtual A {
     active = 1
     address = x.x.x.200 em1:4
     vip_nmask = 255.255.255.0
     port = 1000
     expect = "OK"
     use_regex = 0
     send_program = "/etc/sysconfig/ha/check_udp1000.sh %h"
     load_monitor = none
     scheduler = rr
     protocol = udp
     timeout = 30
     reentry = 15
    quiesce_server = 0
     server A_server1 {
         address = x.x.x.140
         active = 1
         weight = 1
     }
}
virtual B {
     active = 1
     address = x.x.x.202 em1:5
     sorry_server = x.x.x.203
     vip_nmask = 255.255.255.0
     port = 1000
     expect = "OK"
     use_regex = 0
     send_program = "/etc/sysconfig/ha/check_ping.sh %h"
     load_monitor = none
     scheduler = rr
     protocol = udp
     timeout = 30
     reentry = 15
     quiesce_server = 0
     server B_server1 {
         address = x.x.x.146
         active = 1
         weight = 1
     }
     server B_server2 {
         address = x.x.x.147
         active = 1
         weight = 1
     }
}
virtual C {
     active = 1
     address = x.x.x.203 em1:6
     sorry_server = x.x.x.202
     vip_nmask = 255.255.255.0
     port = 1000
     expect = "OK"
     use_regex = 0
     send_program = "/etc/sysconfig/ha/check_udp1000.sh %h"
     load_monitor = none
     scheduler = rr
     protocol = udp
     timeout = 30
     reentry = 15
     quiesce_server = 0
     server C_server1 {
         address = x.x.x.146
         active = 1
         weight = 1
     }
     server C_server2 {
        address = x.x.x.145
         active = 1
         weight = 1
     }
}
virtual D {
     active = 1
     address = x.x.x.201 em1:8
     vip_nmask = 255.255.255.0
     port = 1000
     expect = "OK"
     use_regex = 0
     send_program = "/etc/sysconfig/ha/check_udp1000.sh %h"
     load_monitor = none
     scheduler = rr
     protocol = udp
     timeout = 30
     reentry = 15
     quiesce_server = 0
     server D_server1 {
         address = x.x.x.144
         active = 1
         weight = 1
     }
}

Any idea what could be causing this "limitation" and how I might be able to resolve this?

Thanks,
Will


More information about the lvs-users mailing list