FYI... we traced this back to an OS issue, haven't pin pointed it yet but
another indicator of an issue in the system is that a ping to LOCALHOST or
LOOPBACK (the 127.0.0.1 address) results in the first packet lost and
response times of 1000 ms.

I was working at upgrading the LANFree agent and it was failing on the
deployment engine installation and one of the possible issues there was
that you can't ping your loop back address... I tested that and discovered
the issue mentioned above.

We have seen that when LANFree is running OK, pings to loopback are OK,
when LANFree is running bad, so are the pings to loopback... a reboot of
the system fixes things for about 5 dyas.

At this point I've turned things over to the AIX support staff.

Thought I'd give an update in case anyone runs into this in their
environment.

Dwight E. Cook
Technical Services Prof. Sr.
Strategic Outsourcing Delivery
(918) 493-4678



From:   Dwight Cook <cookde AT COX DOT NET>
To:     ADSM-L AT VM.MARIST DOT EDU
Date:   03/17/2015 02:21 PM
Subject:        TDP/ERP slows to a crawl when only one session remains.
Sent by:        "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>



OK, seeing a strange one.

I have a 10 TB DB

Running 6 sessions with multiplexing of 4

Tdp/erp will run at 300 GB/hr/session lan free until only one session
remains. then the through put drops to about 400 MB/hr

So I'm moving a total of 1700 GB/hr until the last 4 multiplexed files
into
a single session and it grinds to a painfully slow rate.



AIX 5 with tdp/erp is 6.2.1.0  and tsm client 6.2.5.4 going to a 6.3.5.100
server



Has anyone else ever seen anything like this???



And yes, I know all this is old stuff but I deal with what I'm dealt.



Just curious if this has been seen before. I have a sev-2 with level-2
support but they are slow to get back with me.



Thanks in advance,



Dwight