We are suspecting memory leak in our critical production system

Issue

• It seems one of our application server going high memory utilization (above 80%) suspecting there is memory leak.
• We noticed total 32 gb of memory, 11gb allocated for apps and 7gb is free, need to know about 14gb of balance utilization.
• Please note that the unaccounted memory size varies throughout the day. As of now we see approximately 7.93 GiB memory is unaccounted, approximately 13 GB is used by processes, and 2 GB is used by unreclaimable slab.

$ cat sos_commands/memory/free_-m 

              total        used        free      shared  buff/cache   available
Mem:          31884       23889        2542         412        5453        7071   ----------->> Available Memory
Swap:          4095        2234        1861

• Total used RAM on system is '23.79' GiB and the memory consumed by the process is '13.8' GiB:

$ cat ps | gawk '{sum+=$6} END {print "Total RSS : " sum/1024^2 " GiB"}'
Total RSS : 13.8036 GiB

• The top MEM-eating threads are:

    USER      PID    %CPU  %MEM  VSZ-MiB  RSS-MiB  TTY    STAT   START  TIME     COMMAND  
    iloadadm  29637  18.5  16.6  15791    5297     ?      -      Jun22  8601:15  /usr/java/latest/bin/java -DEPOD_MODULE_NAME=ILOAD2 -server 
    iloadadm  24463  9.2   14.6  14565    4678     ?      -      Jun22  4301:18  /usr/java/latest/bin/java -DEPOD_MODULE_NAME=ILOAD2A -server 
    iloadadm  8841   3.2   2.8   10979    919      ?      -      Jul19  227:15   /usr/java/latest/bin/java -DEPOD_MODULE_NAME=ILOAD2SFTP -server 
    iloadadm  15778  0.2   1.2   5110     402      ?      -      02:38  0:32     /usr/java/latest/bin/java -DEPOD_MODULE_NAME=ILOAD2RECO -server 
    root      1772   24.6  1.2   4870     401      ?      -      04:23  39:23    /opt/traps/bin/pmd 
    root      14687  0.0   0.8   1309     277      ?      -      Jun22  0:00     /sbin/vxencryptd -m 
    root      10303  40.3  0.7   473      249      pts/3  -      07:03  0:03     /usr/bin/python /usr/lib/python2.7/site-packages/insights_client/run.py --offline 
    root      5631   44.9  0.4   1168     144      pts/3  -      07:03  0:22     /usr/bin/python /usr/sbin/sosreport 
    root      23820  0.0   0.4   225      133      ?      -      Jun22  29:19    /opt/VRTSvcs/bin/had 
    root      9649   0.7   0.4   1534     129      ?      -      Jun22  364:19   /opt/ds_agent/ds_agent -w /var/opt/ds_agent 

• The slab is around '2.65' GiB:

$ cat proc/meminfo 
Slab:            2659304 kB    --------->>
SReclaimable:     492752 kB
SUnreclaim:      2166552 kB

• So we suspect there is something running on our system which we are not able to catch through the usual means.