I'm hoping someone else has seen this, or could suggest where to look.
On this host
Host PCP CPU Operating System
----------- -------- ------ ------------------
vm03 3.7.1 x86_64 Fedora 18 (Spherical Cow)
configured thusly
kenj@vm03:~$ pmconfig -L
pmapi_version=2
multi_threaded=true
fault_injection=false
secure_sockets=false
ipv6=true
I am seeing a bunch of QA failures, and indeed any remote PCP client command
fails the same way. The symptom is
kenj@bozo $ pmprobe -h vm03 -Dall sample.long.one
__pmSetDataIPC: fd=1024 data=0x7fff51a59450(sz=16)
IPC table fd(PDU version):
__pmSetSocketIPC: fd=1024
IPC table fd(PDU version):
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmDataIPC: fd=1024, data=0x152cd88(sz=16)
__pmConnectPMCD(vm03): pmcd connection port=44321 failed: No route to host
pmNewContext(1, vm03) -> -113, curcontext=-1
pmprobe: Cannot connect to PMCD on host "vm03": No route to host
Now the "No route to host" is totally bogus ...
kenj@bozo $ traceroute vm03
traceroute to vm03 (192.168.1.203), 30 hops max, 60 byte packets
1 vm03.localdomain (192.168.1.203) 0.409 ms !X 0.377 ms !X 0.360 ms !X
kenj@bozo $ ping -c 2 vm03kenj@bozo:~/Logs/by-vm/bozo-laptousingp$ traceroute
vm03
traceroute to vm03 (192.168.1.203), 30 hops max, 60 byte packets
1 vm03.localdomain (192.168.1.203) 0.409 ms !X 0.377 ms !X 0.360 ms !X
kenj@bozo $ ping -c 2 vm03
PING vm03.localdomain (192.168.1.203) 56(84) bytes of data.
64 bytes from vm03.localdomain (192.168.1.203): icmp_req=1 ttl=64 time=0.420 ms
64 bytes from vm03.localdomain (192.168.1.203): icmp_req=2 ttl=64 time=0.362 ms
--- vm03.localdomain ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1001ms
rtt min/avg/max/mdev = 0.362/0.391/0.420/0.029 ms
PING vm03.localdomain (192.168.1.203) 56(84) bytes of data.
64 bytes from vm03.localdomain (192.168.1.203): icmp_req=1 ttl=64 time=0.420 ms
64 bytes from vm03.localdomain (192.168.1.203): icmp_req=2 ttl=64 time=0.362 ms
--- vm03.localdomain ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1001ms
rtt min/avg/max/mdev = 0.362/0.391/0.420/0.029 ms
Apparently it does not matter if the client host is configured to use secure
sockets or not, and similarly for ipv6, as the following examples show:
kenj@bozo $ pmconfig -L
pmapi_version=2
multi_threaded=true
fault_injection=false
secure_sockets=true
ipv6=true
kenj@bozo $ pmprobe -h vm03 sample.long.one
pmprobe: Cannot connect to PMCD on host "vm03": No route to host
kenj@vm16:~$ pmconfig -L
pmapi_version=2
multi_threaded=true
fault_injection=false
secure_sockets=false
ipv6=true
kenj@vm16:~$ pmprobe -h vm03 sample.long.one
pmprobe: Cannot connect to PMCD on host "vm03": No route to host
kenj@vm05:~$ pmconfig -L
pmapi_version=2
multi_threaded=true
fault_injection=false
secure_sockets=false
ipv6=false
kenj@vm05:~$ pmprobe -h vm03 sample.long.one
pmprobe: Cannot connect to PMCD on host "vm03": No route to host
I'm stumped!
|