we are  working in a full lpar copy scenario in a customer environment with a vmware  with 4 CPUs, 32 RAM GB 

In our device map we defined 3   ZDT processors and  25GB of memory .

After the IPL , when we are working via TSO ( even  with just 1 user connected) , we are frequently facing hang situations simply after click the "ENTER" key.

In order to work again we need to disconnect / riconnect the TSO user, but after few seconds of activity still we are facing the same  issue ( hang )

When the system is running the CPU used is  almost 10% , but the memory is used at 90%

Do you have any suggestions? Many thanks in advance

Below you can see more LINUX SERVER DETAILS :

Architecture:          x86_64
CPU op-mode(s):        32-bit, 64-bit
Byte Order:            Little Endian
CPU(s):                4
On-line CPU(s) list:   0-3
Thread(s) per core:    1
Core(s) per socket:    4
CPU socket(s):         1
NUMA node(s):          1
Vendor ID:             GenuineIntel
CPU family:            6
Model:                 15
Stepping:              1
CPU MHz:               2933.437
BogoMIPS:              5866.87
L1d cache:             32K
L1i cache:             32K
L2 cache:              256K
L3 cache:              8192K
NUMA node0 CPU(s):     0-3

Hi Giuseppe, we saw this issue once caused by cron Linux customization.

It's not caused by zD&T or z/OS configuration.

Could you please check the network?

Hi,

Could you check how many I/O devices you are running with? You may need to change the shared memory size and other Linux parameters.

See below inputs if you have more I/O devices

1. In the Devmap you have Memory stanza. It should be more than 5500m (Preferably). Keep it around 10000m. Now, your shrmax will change according. The kernel.shrmmax  will be at-least 1.1 times of the memory defined in the devmap. On the safer side, keep is 1.5 times of memory stanza in the devmap. Use ulimit -m unlimited and ulimit -v unlimited for more than 128 emulated I/O devices
2. kernel.sem  Use as 250 32000 250 1024
3. kernel.msgmni  it should be kernel.msgmni = (350 + 3 * number-of-I/O-devices). Now, check you devmap definition and find out how many input-output devices you have. Now, to find out i/o devices. You will have definition of OSA, 3270 and DASD devices in the devmap. Count of them. In below example. I have 5 i/o devices. So the value will be 350+15

[manager]  # $tapdev define network adapter (OSA) for communication with Linux

name awsosa 0009 --path=A0 --pathtype=OSD --tunnel_intf=y   # QDIO mode

device 7A06 osa osa --unitadd=0

device 7A1A osa osa --unitadd=1

# DASD Devices are defined below. This list was selected during component and application creationg

[manager]

name awsckd 0001

device 2019 3390 3390 /home/ibmsys1/volumes/Z51002

device 201A 3390 3390 /home/ibmsys1/volumes/Z51003

device 2018 3390 3390 /home/ibmsys1/volumes/Z51001

below is the sample of Linux configuration with more than 2000 I/O devices.

Linux configuration can be changed anytime using below methods.

You can issue command  ipcs -l  to see the latest parameters.

You can open the /etc/sysctl.conf file for changes and after changes. then,  run the command sysctl -p to have changes in place.