Good Evening,,

what is the table of basic admin responsibility in maintenance as if I have 50 servers which need to be maintained like what is the expected time to do a routine reboot to have a full refresh and what should I communicate regularly about with the system owners

What time do you reboot boxes?? Never.

We can't define your business processes for you, that's really not something we can help you with. These things aren't technical issues.

Yes we never restart unless if we have a server hung or system crash and that is wrong as far as I know. I checked over internet for administration standereds but i couldn't find any

I want to know for example what is the recommended period to do routine reboot for RedHat particularly?
is there any routine to check already mounted SAN disks? because there is a default disk re-check every 200 days but you have to reboot for that to be activated despite the lack of time going to be spent on that process!

there are alot of general routine procedures to be done by an admin I guess,

I thought there is common standered to ask about!

Again, there should never be a need to reboot. RedHat would NEVER recommend one.

as far as disk scans, you're welcome to unmount and fsck it whenever you see fit. You should never need to though.

There are plenty of standards, but you've not asked about anything that should be standardized.

OK then,, I will wait blindly for a crash or hung then I should pay attention .

blindly? no, you monitor log files etc.

As acid_kewpie said, obviously you should monitor the health of your systems eg disk space, load etc.
Exactly what needs monitoring is really down to how your systems are used.
There are many many tools to do this and you could script your own checks or you could use a pre-built system like Nagios+Cacti; your choice.

Check that backups complete successfully and do test restores every so often to ensure they are not corrupt.

Re reboot: if you install a new kernel, you'll need to reboot to activate it.

If you really want advice, you'd need to describe your setup.

Some admins prefer to reboot when there's a kernel update.

sure (well... not really... you can't "prefer" to reboot for a kernel update (ksplice aside)), but that's a long way away from "8am, every 3rd thursday of the month as per page 37, item 4 of the 12th revision of the Global Corps Ltd. Internal System Operation Manual" isn't it?

I am side tracked now though with the idea of a Systems Admonisher, someone who's role is to tell servers off for misbehaving.

I wasn't saying that he should have a scheduled reboot at certain time intervals. I've seen servers with uptime a little over a year. I've also seen that people want all their software upgraded including the kernel.

Yep, and whichever perspective you **decide** on that all requires thoughts and knowledge, not a list of orders.

The argument you're talking about is not technical but logical/administrative.

Red Hat System Administration needs to consider several non-technical stuff like

1. Planning your BackUp depending your data needs
2. Planning a restore whatever your system crash-down
3. Analysis of company's needs
4. Automating everything
5. Resource management
6. Documentation about your work

And much more. You could give a check about "Red Hat System Administration" , i have got one but it's in italian and for Red Hat 4 so i guess it's useless for you.

Agreed, I wasn't suggesting that he do that.

As far as Linux Admin books/manuals go have a look at http://www.linuxtopia.org/online_boo...ion_index.html, but they are mostly about install and setup, not ongoing tasks.