To check total disk usage and inodes usage, please run following commands: (note, that we assume xtm is installed in its default location)

du -sch /xtm /servers /var/lib/pgsql/ 

And watch for combined disk usage. A sample below:

16G /servers
5.9G /xtm
85M /var/lib/pgsql/
22G total

Now, we can check for inodes in the same way:

du -sch --inodes /servers /xtm /var/lib/pgsql/

And a result of it, showing total inodes used by these directories:

13K /servers
47K /xtm
2.6K /var/lib/pgsql/
62K total


A most basic check is to run the command:

top -u xtm

and then check the "RES" column – which shows the "Resident size" for processes run by xtm user ( the application runs under xtm user privileges )

A more comprehensive way to check how much memory is used by xtm application can be accomplished by running the command below:

smem -tp -P java -c "name pss"

Sample output:

Command                     PSS
python /bin/smem -tp -P jav 0.16%
/usr/bin/java -Xms64M -Xmx1 8.23%
/usr/bin/java -Xms64M -Xmx1 8.60%

In this example, we use a percentage instead of numeric values and we narrow down the list to only show the java processes. You can tailor it to your own needs with many switches – please refer to man smem. Here we also showed the main component usage, for total usage, please read the note below:

NOTE: To get actual usage, you need also include not only java processes, but also e.g. httpd service, as it also runs simultaneously with the application, and then sum up all the values.


As with the memory example, to count total cpu usage we need to sum up a %of CPU time of all components of xtm: the java processes, user processes and httpd processes.

 It can be achieved by running top command:

PID     USER    PR NI   VIRT    RES     SHR     S %CPU  %MEM    TIME+   COMMAND
4723    xtm     20 0    4573008 331828  16596   0.7   8.6     0:33.90 java
10485   root    20 0    61848   2320    1584    0.7   0.1     0:00.25 top
4722    xtm     20 0    4578148 340276  16604   0.6   8.8     0:38.07 java

There you need to sum up the %cpu of java, httpd and additional user's processes to get the grand total value.