FlexLM and pgroupd IPC

Accelerated Computing HPC Compilers Legacy PGI Compilers
#1

Hi PGI Forum,

I have tried the suggest license with our PGI/Flexlm pgroupd. I seem to be having problems with the IPC between flexlm and pgroupd.

I receive the following errors (I tried both by changing the license file on the DAEMON LINE):

pgroupd in linux86/bin:

12:13:12 (lmgrd) pgroupd using TCP-port 33132
12:13:15 (pgroupd) Server unable to setup IPC channel
12:13:15 (pgroupd) EXITING DUE TO SIGNAL 39 Exit reason 7
12:13:15 (lmgrd) pgroupd exited with status 39 (Couldn’t find a master)
12:13:15 (lmgrd) REStarted pgroupd (internet tcp_port 38144 pid 8800)
12:13:15 (pgroupd) FLEXnet Licensing version v11.5.0.0 build 56285 i86_se9
12:13:15 (pgroupd) Server started on falcon.ansto.gov.au for:
12:13:15 (pgroupd) pgi72-server-complete (consisting of: pgi-hpf-lin64

12:13:15 (pgroupd) pgi-f95-lin64 pgi-f77-lin64 pgi-cc-lin64
12:13:15 (pgroupd) pgi-cpp-lin64 pgi-hpf-lin32 pgi-f95-lin32
12:13:15 (pgroupd) pgi-f77-lin32 pgi-cc-lin32 pgi-cpp-lin32
12:13:15 (pgroupd) pgi-prof pgi-dbg pgi-dbg-gui)
12:13:15 (pgroupd) pgi71-server-complete (consisting of: pghpf-linux86
-64
12:13:15 (pgroupd) pgf90-linux86-64 pgf77-linux86-64 pgcc-linux86-64
12:13:15 (pgroupd) pgcpp-linux86-64 pghpf-linux86 pgf90-linux86
12:13:15 (pgroupd) pgf77-linux86 pgcc-linux86 pgcpp-linux86
12:13:15 (pgroupd) pgprof pgdbg pgdbg-gui
12:13:15 (pgroupd) pgdbg-linux86-64) pghpf-linux86-64 pgf90-linux86-64
12:13:15 (pgroupd) pgf77-linux86-64 pgcc-linux86-64 pgcpp-linux86-64
12:13:15 (pgroupd) pghpf-linux86 pgf90-linux86 pgf77-linux86
12:13:15 (pgroupd) pgcc-linux86 pgcpp-linux86 pgprof
12:13:15 (pgroupd) pgdbg pgdbg-gui pgdbg-linux86-64
12:13:15 (pgroupd) pgi-hpf-lin64 pgi-f95-lin64 pgi-f77-lin64
12:13:15 (pgroupd) pgi-cc-lin64 pgi-cpp-lin64 pgi-hpf-lin32
12:13:15 (pgroupd) pgi-f95-lin32 pgi-f77-lin32 pgi-cc-lin32
12:13:15 (pgroupd) pgi-cpp-lin32 pgi-prof pgi-dbg
12:13:15 (pgroupd) pgi-dbg-gui
12:13:15 (pgroupd) EXTERNAL FILTERS are OFF
12:13:15 (lmgrd) pgroupd using TCP-port 38144
12:13:15 (pgroupd) Server unable to setup IPC channel
12:13:15 (pgroupd) EXITING DUE TO SIGNAL 39 Exit reason 7
12:13:15 (lmgrd) pgroupd exited with status 39 (Couldn’t find a master)
12:13:15 (lmgrd) REStarted pgroupd (internet tcp_port 39465 pid 8803)

– > And this keeps looping trying different random ports for the IPC.

pgroupd in linux86-64/bin:

12:06:59 (lmgrd) manager (lmgrd) will attempt to re-start the vendor daemon.
12:06:59 (lmgrd) REStarted pgroupd (internet tcp_port 44748 pid 8174)
12:06:59 (pgroupd) FLEXnet Licensing version v11.5.0.0 build 56285 x64_se9
12:06:59 (pgroupd) Server started on falcon.ansto.gov.au for:
12:06:59 (pgroupd) pgi72-server-complete (consisting of: pgi-hpf-lin64

12:06:59 (pgroupd) pgi-f95-lin64 pgi-f77-lin64 pgi-cc-lin64
12:06:59 (pgroupd) pgi-cpp-lin64 pgi-hpf-lin32 pgi-f95-lin32
12:06:59 (pgroupd) pgi-f77-lin32 pgi-cc-lin32 pgi-cpp-lin32
12:06:59 (pgroupd) pgi-prof pgi-dbg pgi-dbg-gui)
12:06:59 (pgroupd) pgi71-server-complete (consisting of: pghpf-linux86
-64
12:06:59 (pgroupd) pgf90-linux86-64 pgf77-linux86-64 pgcc-linux86-64
12:06:59 (pgroupd) pgcpp-linux86-64 pghpf-linux86 pgf90-linux86
12:06:59 (pgroupd) pgf77-linux86 pgcc-linux86 pgcpp-linux86
12:06:59 (pgroupd) pgprof pgdbg pgdbg-gui
12:06:59 (pgroupd) pgf77-linux86-64 pgcc-linux86-64 pgcpp-linux86-64
12:06:59 (pgroupd) pghpf-linux86 pgf90-linux86 pgf77-linux86
12:06:59 (pgroupd) pgcc-linux86 pgcpp-linux86 pgprof
12:06:59 (pgroupd) pgdbg pgdbg-gui pgdbg-linux86-64
12:06:59 (pgroupd) pgi-hpf-lin64 pgi-f95-lin64 pgi-f77-lin64
12:06:59 (pgroupd) pgi-cc-lin64 pgi-cpp-lin64 pgi-hpf-lin32
12:06:59 (pgroupd) pgi-f95-lin32 pgi-f77-lin32 pgi-cc-lin32
12:06:59 (pgroupd) pgi-cpp-lin32 pgi-prof pgi-dbg
12:06:59 (pgroupd) pgi-dbg-gui
12:06:59 (pgroupd) EXTERNAL FILTERS are OFF
12:06:59 (lmgrd) pgroupd using TCP-port 44748
*** glibc detected *** pgroupd: free(): invalid pointer: 0x00007fff674e7d10 ***
======= Backtrace: =========
/lib64/libc.so.6[0x332e671684]
/lib64/libc.so.6(cfree+0x8c)[0x332e674ccc]
pgroupd[0x40f9b0]
pgroupd[0x40f739]
/lib64/libc.so.6(__libc_start_main+0xf4)[0x332e61d8b4]
pgroupd(tolower+0x42)[0x4032fa]
======= Memory map: ========
00400000-00530000 r-xp 00000000 fd:02 688186 /opt/pgi/lin
ux86-64/7.2-5/bin/pgroupd
00630000-00635000 rw-p 00130000 fd:02 688186 /opt/pgi/lin
ux86-64/7.2-5/bin/pgroupd
00635000-00642000 rw-p 00635000 00:00 0
00f67000-00fa1000 rw-p 00f67000 00:00 0
332e200000-332e21a000 r-xp 00000000 08:01 65838 /lib64/ld-2.
5.so
332e41a000-332e41b000 r–p 0001a000 08:01 65838 /lib64/ld-2.
5.so
332e41b000-332e41c000 rw-p 0001b000 08:01 65838 /lib64/ld-2.
5.so
332e600000-332e74a000 r-xp 00000000 08:01 65839 /lib64/libc-
2.5.so
332e74a000-332e949000 —p 0014a000 08:01 65839 /lib64/libc-
2.5.so
332e949000-332e94d000 r–p 00149000 08:01 65839 /lib64/libc-
2.5.so
332e94d000-332e94e000 rw-p 0014d000 08:01 65839 /lib64/libc-
2.5.so
332e94e000-332e953000 rw-p 332e94e000 00:00 0
332ee00000-332ee02000 r-xp 00000000 08:01 65845 /lib64/libdl
-2.5.so
332ee02000-332f002000 —p 00002000 08:01 65845 /lib64/libdl
-2.5.so
332f002000-332f003000 r–p 00002000 08:01 65845 /lib64/libdl
-2.5.so
332f003000-332f004000 rw-p 00003000 08:01 65845 /lib64/libdl
-2.5.so
332f200000-33
332f215000-332f414000 —p 00015000 08:01 65840 /lib64/libpt
hread-2.5.so
332f414000-332f415000 r–p 00014000 08:01 65840 /lib64/libpt
hread-2.5.so
332f415000-332f416000 rw-p 00015000 08:01 65840 /lib64/libpt
hread-2.5.so
332f416000-332f41a000 rw-p 332f416000 00:00 0
3337000000-333700d000 r-xp 00000000 08:01 65746 /lib64/libgc
c_s-4.1.2-20080102.so.1
333700d000-333720d000 —p 0000d000 08:01 65746 /lib64/libgc
c_s-4.1.2-20080102.so.1
333720d000-333720e000 rw-p 0000d000 08:01 65746 /lib64/libgc
c_s-4.1.2-20080102.so.1
3338800000-3338811000 r-xp 00000000 08:01 65851 /lib64/libre
solv-2.5.so
3338811000-3338a11000 —p 00011000 08:01 65851 /lib64/libre
solv-2.5.so
3338a11000-3338a12000 r–p 00011000 08:01 65851 /lib64/libre
solv-2.5.so
3338a12000-3338a13000 rw-p 00012000 08:01 65851 /lib64/libre
solv-2.5.so
3338a13000-3338a15000 rw-p 3338a13000 00:00 0
2b8e435c1000-2b8e435c3000 rw-p 2b8e435c1000 00:00 0
2b8e435db000-2b8e435de000 rw-p 2b8e435db000 00:00 0
2b8e435f7000-2b8e43601000 r-xp 00000000 08:01 65564 /lib64/libns
s_files-2.5.so
2b8e43601000-2b8e43800000 —p 0000a000 08:01 65564 /lib64/libns
s_files-2.5.so
2b8e43800000-2b8e43801000 r–p 00009000 08:01 65564 /lib64/libns
s_files-2.5.so
2b8e43801000-2b8e43802000 rw-p 0000a000 08:01 65564 /lib64/libns
s_files-2.5.so
2b8e43802000-2b8e43806000 r-xp 00000000 08:01 65562 /lib64/libns
s_dns-2.5.so
2b8e43806000-2b8e43a05000 —p 00004000 08:01 65562 /lib64/libns
s_dns-2.5.so
2b8e43a05000-2b8e43a06000 r–p 00003000 08:01 65562 /lib64/libns
s_dns-2.5.so
2b8e43a06000-2b8e43a07000 rw-p 00004000 08:01 65562 /lib64/libns
s_dns-2.5.so
2b8e44000000-2b8e44021000 rw-p 2b8e44000000 00:00 0
2b8e44021000-2b8e48000000 —p 2b8e44021000 00:00 0
7fff674d4000-7fff674e9000 rw-p 7fff674d4000 00:00 0 [stack]
ffffffffff600000-ffffffffffe00000 —p 00000000 00:00 0 [vdso]
12:07:02 (lmgrd) pgroupd exited with status 0 signal = 17
12:07:02 (lmgrd) Since this is an unknown status, license server
12:07:02 (lmgrd) manager (lmgrd) will attempt to re-start the vendor daemon.
12:07:02 (lmgrd) REStarted pgroupd (internet tcp_port 45395 pid 8178)

Please let me know if you require any more data here.

Regards
Cooper Lees

#2

You are using the correct pgroupd daemon (version 11.5), so we
suspect the firewall is causing the failures.

First make sure that the you can
ping xxxxx
from the platform running the compilers,
where xxxxx is the hostname in the SERVER line of the
license file. If not, edit the file, or change permissions
and /etc/hosts file to allow that to occur.

Also make sure that
$PGI/linux86/7.2/bin/lmutil lmhostid
output matches the hostid in the SERVER line of the license file.


If everything checks out, try turning off the firewall and see if the
license works. If it does, then you should make sure that
lmgrd is allowed to use PORT 27000 (see SERVER line of license)
by the firewall, and you may also need to assign a PORT to pgroupd
and tell the firewall to allow it as well.

change the line
DAEMON pgroupd
to
DAEMON pgroupd PORT=1234
and pgroupd will always use PORT 1234 for its use.

#3

I can confirm there is no iptables running on falcon, so it is very weird a process running on the same machine can not bind to a socket for the IPC.

[root@falcon pgi]# pgroupd -v
10:30:01 (pgroupd) FLEXnet Licensing version v11.5.0.0 build 56285 x64_se9
pgroupd v11.5.0.0 build 56285 x64_se9 - Copyright © 1988-2007 Macrovision Europe Ltd. and/or Macrovision Corporation. All Rights Reserved.
[root@falcon pgi]# /etc/init.d/iptables status
Firewall is stopped.
[root@falcon pgi]# ping falcon
PING falcon.ansto.gov.au (x) 56(84) bytes of data.
64 bytes from falcon.ansto.gov.au (x): icmp_seq=1 ttl=64 time=0.023 ms
64 bytes from falcon.ansto.gov.au (x): icmp_seq=2 ttl=64 time=0.005 ms
64 bytes from falcon.ansto.gov.au (x): icmp_seq=3 ttl=64 time=0.004 ms
64 bytes from falcon.ansto.gov.au (x): icmp_seq=4 ttl=64 time=0.003 ms

— falcon.ansto.gov.au ping statistics —
4 packets transmitted, 4 received, 0% packet loss, time 2999ms
rtt min/avg/max/mdev = 0.003/0.008/0.023/0.009 ms
[root@falcon pgi]# $PGI/linux86/7.2/bin/lmutil lmhostid
lmutil - Copyright © 1989-2007 Macrovision Europe Ltd. and/or Macrovision Corporation. All Rights Reserved.
The FLEXnet host ID of this machine is “001e685782d8”
—> This matches the ID in my license file …

I did define a static port like you suggested and still the same error but it now states the port I assigned, as one would expect:

Log File:

10:15:55 (pgroupd) EXTERNAL FILTERS are OFF
10:15:55 (lmgrd) pgroupd using TCP-port 12345
10:15:55 (pgroupd) Server unable to setup IPC channel
10:15:55 (pgroupd) EXITING DUE TO SIGNAL 39 Exit reason 7
10:15:55 (lmgrd) pgroupd exited with status 39 (Couldn’t find a master)
10:15:55 (lmgrd) Using vendor daemon port 12345 specified in license file
10:15:55 (lmgrd) REStarted pgroupd (internet tcp_port 12345 pid 10411)
10:15:55 (pgroupd) FLEXnet Licensing version v11.5.0.0 build 56285 i86_se9

What else can cause the IPC to fail? What does “status 39” and “reason 7” mean?

#4

Hi all,

I have solved this issue. It was my screw up. ALWAYS make sure you /etc/hosts file is not screwed. Sorry about wasting your time PG.