Hi condor users,
I am using Condor Version 7.6.6 in a
pool of Windows desktops, and I have a NIC binding issue when the Desktop
has more then one NIC. We also used before the v7.2.4 and the 7.4.x series.
In a particular desktop (PC315286, Windows
XP Pro x64 Edition), we have two physical NIC e two VMware instances NICs.
It always binds to a NIC that has no
cable connected and when I disable it, it binds to one of the VMware NICs.
See Ipconfig /all output below:
Windows IP Configuration
Host Name . . . . . . . . . . . . : PC315286
Primary Dns Suffix . . . . . .
. : domain1
Node Type . . . . . . . . . . . . : Hybrid
IP Routing Enabled. . . . . . . . : No
WINS Proxy Enabled. . . . . . . . : No
DNS Suffix Search List. . . . . . : domain1
domain2
domain3
Ethernet adapter VMware Network Adapter VMnet8:
Connection-specific DNS Suffix .
:
Description . . . . . . . . . . . : VMware
Virtual Ethernet Adapter for VMnet8
Physical Address. . . . . . . . . : 00-50-56-C0-00-08
DHCP Enabled. . . . . . . . . . . : No
IP Address. . . . . . . . . . . . : 192.168.179.1
Subnet Mask . . . . . . . . . . . : 255.255.255.0
Default Gateway . . . . . . . . . :
Ethernet adapter VMware Network Adapter VMnet1:
Connection-specific DNS Suffix .
:
Description . . . . . . . . . . . : VMware
Virtual Ethernet Adapter for VMnet1
Physical Address. . . . . . . . . : 00-50-56-C0-00-01
DHCP Enabled. . . . . . . . . . . : No
IP Address. . . . . . . . . . . . : 192.168.206.1
Subnet Mask . . . . . . . . . . . : 255.255.255.0
Default Gateway . . . . . . . . . :
Ethernet adapter Local Area Connection 2:
Connection-specific DNS Suffix .
:
Description . . . . . . . . . . . : Realtek
RTL8169/8110 Family Gigabit Ethernet NIC
Physical Address. . . . . . . . . : 00-1A-3F-7B-67-F3
DHCP Enabled. . . . . . . . . . . : Yes
Autoconfiguration Enabled . . . . : Yes
Autoconfiguration IP Address. . . : 169.254.188.36
Subnet Mask . . . . . . . . . . . : 255.255.0.0
Default Gateway . . . . . . . . . :
Ethernet adapter Local Area Connection:
Connection-specific DNS Suffix .
: domain2
Description . . . . . . . . . . . : Intel(R)
82567LM-3 Gigabit Network Connection
Physical Address. . . . . . . . . : 3C-4A-92-CE-5E-3D
DHCP Enabled. . . . . . . . . . . : Yes
Autoconfiguration Enabled . . . . : Yes
IP Address. . . . . . . . . . . . : 10.3.28.14
Subnet Mask . . . . . . . . . . . : 255.255.254.0
Default Gateway . . . . . . . . . : 10.3.29.254
DHCP Server . . . . . . . . . . . : 10.x.y.142
DNS Servers . . . . . . . . . . . : 10.w.z.246
10.x.y.141
Primary WINS Server . . . . . . . : 10.x.y.18
Secondary WINS Server . . . . . . : 10.x.y.11
Lease Obtained. . . . . . . . . . : quarta-feira,
1 de agosto de 2012 11:54:56
Lease Expires . . . . . . . . . . : quinta-feira,
2 de agosto de 2012 11:54:56
The network Condor should bind to is
"Local Area Connection"
above!
The MasterLog shows:
08/01/12 11:21:17 DaemonCore: command socket at <169.254.127.217:24512>
08/01/12 11:21:17 DaemonCore: private command socket
at <169.254.127.217:24512>
08/01/12 11:21:17 Setting maximum accepts per cycle
4.
08/01/12 11:21:18 Started DaemonCore process "C:\condor/bin/condor_schedd.exe",
pid and pgroup = 3232
08/01/12 11:21:19 Started DaemonCore process "C:\condor/bin/condor_startd.exe",
pid and pgroup = 3252
08/01/12 11:21:19 Started DaemonCore process "C:\condor/bin/condor_kbdd.exe",
pid and pgroup = 3260
08/01/12 11:21:30 PERMISSION DENIED to unauthenticated@unmapped
from host 169.254.127.217 for command 60008 (DC_CHILDALIVE), access level
DAEMON: reason: DAEMON authorization policy contains no matching ALLOW
entry for this request; identifiers used for this host: 169.254.127.217,PC315286.corp.ad.emb
Using BIND_ALL_INTERFACES
= TRUE gives the same issue!
When I disable the second NIC, it shows:
08/01/12 11:31:32 DaemonCore: command socket at <192.168.206.1:61402>
08/01/12 11:31:32 DaemonCore: private command socket
at <192.168.206.1:61402>
08/01/12 11:31:32 Setting maximum accepts per cycle
4.
08/01/12 11:31:33 Started DaemonCore process "C:\condor/bin/condor_schedd.exe",
pid and pgroup = 3264
08/01/12 11:31:34 Started DaemonCore process "C:\condor/bin/condor_startd.exe",
pid and pgroup = 3288
08/01/12 11:31:34 Started DaemonCore process "C:\condor/bin/condor_kbdd.exe",
pid and pgroup = 3296
08/01/12 11:31:48 PERMISSION DENIED to unauthenticated@unmapped
from host 192.168.206.1 for command 60008 (DC_CHILDALIVE), access level
DAEMON: reason: DAEMON authorization policy contains no matching ALLOW
entry for this request; identifiers used for this host: 192.168.206.1,PC315286.corp.ad.emb
When I set NETWORK_INTERFACE
= 10.3.28.14 :
08/01/12 11:56:43 DaemonCore: command socket at <10.3.28.14:38946>
08/01/12 11:56:43 DaemonCore: private command socket
at <10.3.28.14:38946>
08/01/12 11:56:43 Setting maximum accepts per cycle
4.
08/01/12 11:56:44 Started DaemonCore process "C:\condor/bin/condor_schedd.exe",
pid and pgroup = 3208
08/01/12 11:56:45 Started DaemonCore process "C:\condor/bin/condor_startd.exe",
pid and pgroup = 3220
08/01/12 11:56:46 Started DaemonCore process "C:\condor/bin/condor_kbdd.exe",
pid and pgroup = 3264
it fixes the issue, but the IP addreses
are dynamic IP and if it changes I need to hardcode the new addres to the
config file.
Prior to version 7.6.6, version 7.2.4
as an example, I used to rearrange the order of the NICs in the "Network
Connections" panel, "Advanced" Menu option, "Advanced
Settings...", under Connections, and Condor always respected this
settings. In v7.2.4, this avoids the use of the NETWORK_INTERFACE
config setting.
In version 7.6.6 Condor does not respect
this settings.
Was this behavior changed from v7.2.4
to v7.6.6? Could someone point in which stable release this was changed?
Is there any other way to overcome this
issue?
Try setting this in your config file: NETWORK_INTERFACE = 10.*
This should cause Condor to use the network interface whose IP address matches the pattern.
Thanks and regards, Jaime Frey UW-Madison Condor Team
|
![[Computer Systems Lab]](https://www-auth.cs.wisc.edu/pics/csl_logo.gif){kind=logo}