Operating System - HP-UX
1747980 Members
4027 Online
108756 Solutions
New Discussion

Re: error IPC0107 UNIX0227

 
NDO
Super Advisor

error IPC0107 UNIX0227

Hi

 

I have two systems running 11.31, running an application of mediation (telecom), in which they run as master & slave, but the application vendor has disable the second node, because of the error above in the subject (IPC0107 UNIX0227),.

So I was asked by the application vendor to suggest that could an error related to socket/network issue, and they ask me to filter all UDP packets in the range 20000 to 20405.

I have run the following command:

tcpdump -vv -X -i lan3 udp portrange 20000-20405

and the output was:

tcpdump: listening on lan3, link-type EN10MB (Ethernet), capture size 65535 bytes
17:26:02.040504 IP (tos 0x0, ttl 1, id 23017, offset 0, flags [DF], proto UDP (17), length 192)
    prep01.51070 > 10.100.48.255.20350: [udp sum ok] UDP, length 164
        0x0000:  4500 00c0 59e9 4000 0111 a972 0a64 300b  E...Y.@....r.d0.
        0x0010:  0a64 30ff c77e 4f7e 00ac 32fd 4801 1084  .d0..~O~..2.H...
        0x0020:  0002 a400 0000 0175 696f 5f6d 6169 6e74  .......uio_maint
        0x0030:  5f6c 6f67 5f73 7276 0000 0000 0000 0000  _log_srv........
        0x0040:  0000 0000 0000 0000 406e 0000 7072 6570  ........@n..prep
        0x0050:  3031 0000 0000 0000 0000 0000 0000 0000  01..............
        0x0060:  0000 0000 0000 0000 0000 0000 0000 0000  ................
        0x0070:  0000 0000 0000 0000 0000 0000 0000 0000  ................
        0x0080:  0000 0000 0000 0000 0000 0000 6367 6900  ............cgi.
        0x0090:  0000 0000 0000 0000 0500 0000 e45e 5d56  .............^]V
        0x00a0:  7072 6570 3031 3a34 3330 4545 3235 3739  prep01:430EE2579
        0x00b0:  3246 3234 3036 3042 3036 3130 3530 3000  2F24060B0610500.
17:26:03.039247 IP (tos 0x0, ttl 1, id 23018, offset 0, flags [DF], proto UDP (17), length 192)
    prep01.51044 > 10.100.48.255.20350: [udp sum ok] UDP, length 164
        0x0000:  4500 00c0 59ea 4000 0111 a971 0a64 300b  E...Y.@....q.d0.
        0x0010:  0a64 30ff c764 4f7e 00ac 374a 4801 1084  .d0..dO~..7JH...
        0x0020:  0002 a400 0000 0164 6973 6b5f 6d67 6d74  .......disk_mgmt
        0x0030:  5f6d 6f6e 6974 6f72 0000 0000 0000 0000  _monitor........
        0x0040:  0000 0000 0000 0000 6055 0000 7072 6570  ........`U..prep
        0x0050:  3031 0000 0000 0000 0000 0000 0000 0000  01..............
        0x0060:  0000 0000 0000 0000 0000 0000 0000 0000  ................
        0x0070:  0000 0000 0000 0000 0000 0000 0000 0000  ................
        0x0080:  0000 0000 0000 0000 0000 0000 6367 6900  ............cgi.
        0x0090:  0000 0000 0000 0000 0500 0000 a25c 5d56  .............\]V
        0x00a0:  7072 6570 3031 3a34 3330 4545 3235 3739  prep01:430EE2579
        0x00b0:  3246 3234 3036 3042 3036 3130 3530 3000  2F24060B0610500.
17:26:03.041677 IP (tos 0x0, ttl 1, id 23019, offset 0, flags [DF], proto UDP (17), length 192)
    prep01.51046 > 10.100.48.255.20350: [udp sum ok] UDP, length 164
        0x0000:  4500 00c0 59eb 4000 0111 a970 0a64 300b  E...Y.@....p.d0.
        0x0010:  0a64 30ff c766 4f7e 00ac f3ad 4801 1084  .d0..fO~....H...
        0x0020:  0002 a400 0000 0174 6173 6b5f 6d61 6e61  .......task_mana
        0x0030:  6765 7200 0000 0000 0000 0000 0000 0000  ger.............
        0x0040:  0000 0000 0000 0000 7955 0000 7072 6570  ........yU..prep
        0x0050:  3031 0000 0000 0000 0000 0000 0000 0000  01..............
        0x0060:  0000 0000 0000 0000 0000 0000 0000 0000  ................
        0x0070:  0000 0000 0000 0000 0000 0000 0000 0000  ................
        0x0080:  0000 0000 0000 0000 0000 0000 6367 6900  ............cgi.
        0x0090:  0000 0000 0000 0000 0500 0000 a25c 5d56  .............\]V
        0x00a0:  7072 6570 3031 3a34 3330 4545 3235 3739  prep01:430EE2579
        0x00b0:  3246 3234 3036 3042 3036 3130 3530 3000  2F24060B0610500.

3 packets captured
304 packets received by filter
0 packets dropped by kernel

But I am unable to interpret this output, can someone help on this?

6 REPLIES 6
Laurent Menase
Honored Contributor

Re: error IPC0107 UNIX0227

Hi,

Probably only the vendor can tell you something about it.

Now to collect the trace you should use the option -w somefile.pcap, and analyze them under wireshark, it will give a more comprehensive view of the packets.

Best regards,

Laurent

 

Akio Kabutogi
Advisor

Re: error IPC0107 UNIX0227

The application vendor means that the application would not work well if some other
application is using the port range 20000-20405 ? If I translate UNIX0277 to unix
errno=227, then it is EADDRNOTAVAIL. The trace shows the system "prep01" is
sending broadcast packet to 10.100.48.255.20350 periodically. If this "prep01" is
the host running tcpdump, then, some process on the local host is sending OUT
those broadcast packets. If this "prep01" is a certain remote host, then, the local
host is receiving the broadcast packet. If it's the latter case, you may want to know if
there's receiver process running on it. If it is running, then, that process should have
port 20350 bound to its own socket. This socket bound to port 20350 may be
causing the issue for the application of mediation (telecom), leading to UNIX0277.

kab
I work for HPE
NDO
Super Advisor

Re: error IPC0107 UNIX0227

Hi

 

Thank you for your responses, host prep01 is the master server of the application, and prep02 (where I run the tcpdump command) is the slave server, where the application vendor has now disabled the application, but O.S. is still UP. I will enquire with app vendor what process is tied to port 20350!

Dennis Handly
Acclaimed Contributor

Re: error IPC0107 UNIX0227

>I will enquire with app vendor what process is tied to port 20350

 

There should be a HP-UX command that will give you the PID so you can cross check.

netstat or lsof?

 

 

NDO
Super Advisor

Re: error IPC0107 UNIX0227

 

netstat -a | grep 20350

Does not give me any output

Akio Kabutogi
Advisor

Re: error IPC0107 UNIX0227

So, prep01(master server on which the application is stopped) is sending broadcast UDP
packet to port 20350. This means prep02 does not need to have port 20350 open. The
source port on prep01 for these UDP packets are 51070, 51044 and 51046. You may find
those ports(or some other ports around that range) in netstat -an|grep udp on prep01.
However, if those messages are the one that would be received also on prep02, then,
you'll find 20350 in netstat -an|grep 20350 on prep02. But they're broadcast packet.
Thus, prep02 does not always need to receive those broadcast packets. In that case,
prep02 does not have port 20350 open and the system just discard them. Probably,
those UDP broadcast packets are received and processed by other system on the same
subnet(10.100.48.xx).

kab
I work for HPE