EMS Monitors Use of TCP Ports -- Is there a defined RANGE? Our APPS are contenting with Ports!

[Alzhy]

I've an outstanding case with HP, but I thought I'd go this route too.

We've a massive apps that require and use a number of ports. The Apps use "listner" processes that transacts with Weblogic ecosystems. The system is still growing and we're poised to add more blocks of tcp port definitions in /etc/services probably all the way up to 65000+. The ports we use start in the 50000'ish.

Lately though, we've instances when our APPS listener processes cannot start or spawn connections becuase the tcpp ports we defined is reported as in USE. And true enough -- it is and those EMS monitors do use them. What's puzzling is the ports in use are not consistent accross systems -- but they are in the 50000'ish range too.

Example from LSOF:

dm_memory 2888 root 5u inet 0x1027d4680 0t198821 TCP 127.0.0.1:50485->127.0.0.1:49980 (ESTABLISHED)
dm_memory 2888 root 7u inet 0x10736a0c0 0t0 TCP *:50110 (LISTEN)
dm_stape 3556 root 5u inet 0xb6ef17c0 0t421537 TCP 127.0.0.1:50486->127.0.0.1:49980 (ESTABLISHED)
dm_FCMS_a 25222 root 2u inet 0x9ebe4ac0 0t0 TCP *:50022 (LISTEN)
dm_FCMS_a 25222 root 7u inet 0x1070ea280 0t268749 TCP 127.0.0.1:50498->127.0.0.1:49980 (ESTABLISHED)
dm_TL_ada 26206 root 2u inet 0x10365f800 0t0 TCP *:50051 (LISTEN)
dm_TL_ada 26206 root 7u inet 0x106003540 0t3319687 TCP 127.0.0.1:50492->127.0.0.1:49980 (ESTABLISHED)
dm_core_h 29948 root 2u inet 0x1060c6700 0t0 TCP *:50090 (LISTEN)
dm_core_h 29948 root 7u inet 0x1070e33c0 0t48246 TCP 127.0.0.1:50497->127.0.0.1:49980 (ESTABLISHED)

[Shinji Teragaito_1]

I guess the following ITRC document may resolv your problem:

http://www11.itrc.hp.com/service/cki/docDisplay.do?docLocale=en&docId=emr_na-c00929065-3

Shinji