Need clarification for the below queries..

[anantham_1]

Dear Gurus,

I am having the below mentioned doubts.Please explain me in detail.

1.I want to know when,whom,from where loged in to server,I need for root as well?
2.Files changed/created since morning 9.00AM
3.From OS how can I check memory slots status.Memory installed or free,If installed what is the size of the memory?
4.Importance of paging and where it is used?
5.SAP installed and already data blocks are created.I want to know how much sapce is free or how much filled in that?

Thanks in advance for your reply.

With Regds,
Anand.

[James R. Ferguson]

Hi Anand:

1. See the 'last(1)' manpages for more information. Use:

# last -R

...If you don't have a '/var/adm/wtmp' file, create one:

# touch /var/adm/wtmp

...the presence of this file starts tracking last login information.

2. To find files created since 0900 today:

# touch -mt 06180900 /tmp/myref
# find /path -xdev -newer /tmp/myref

...This creates a "reference" file whose modification time is June 18 (0618) at 0900. The 'find' command then finds all files whose *modification* timestamp is more recent. There is *no* *creation* timestamp maintained by Unix. A file's creation time can be known only if it has never been modified, since the 'mtime' attribute represents the last modified time. See the manpages for 'touch(1)', 'find(1)' and 'stat(2)' for more information.

3. Use 'cstm' (a part of the Online Diagnostics software), thusly:

# # echo "selclass qualifier memory;info;wait;infolog"|cstm > /tmp/meminfo

4. "Paging" is another (more modern) term for "swapping". Paging allows virtual memory to exceed the real, physical memory on a server by allowing portions of a process's memory to be moved to disk when it is not needed and recalled when it is. In this fashion, more processes can run concurrently than if every process had to remain in the finite physical memory space for the duration the process's life.

5. I am not familar with SAP, so I cannot comment here.

Regards!

...JRF...

[Michael Steele_2]

To track root logins 'su' has to be used and the .profile file updated something like this:

HISTFILE=/.hist/${HOME}/`who am i|awk '{ print $1}'`-`date +%d:%m:%y.%H:%M`
export HISTFILE
HISTSIZE=500
export HISTSIZE

/usr/adm/sulog

su

http://www.rbi.informatik.uni-frankfurt.de/doc/man/hpux/su.1.html