How to Maintain your Linux Environment, if you're New to Linux
Novell Cool Solutions: Feature
By Hanny Kraa
Digg This -
Posted: 16 Aug 2007
When for the first time I had to administer a complete Novell system which had been migrated to Linux, I didn't know what to do. All of my routines were gone and I had to find new ones. I frantically looked over all kinds of pages on the Novell site, which luckily enough were a great help.
But... you have to browse a lot of different pages, because each of them discusses another single subject of the Linux environment.
There is one page which discusses all kinds of different commands, other pages which discuss GroupWise on Linux, and again other pages which describe how to check the health of a Linux server.
I put together some basic information to make my own 'quick reference' for basic Linux administration tasks.
I think it might come in handy for a lot of NetWare sys engineers and admins who are Linux newbies.
Maintain your Linux Environment, for NetWare Administrators who are new to Linux
The specialist team marched in, took away your dearest NetWare servers and migrated the whole system to Linux, just like that.
Or maybe you actually performed the migration all by yourself.
You realize what all the benefits are of migrating to Linux and you feel o so lucky with the brand new Linux environment which you're going to administer.
But.... now it's post-migration day. And you need to do your 'things' just as you did them in your NetWare environment.
Maybe you already know some Linux stuff or maybe you're the jedi master of Linux himself, but even then you'll still discover that you have a vast range of NetWare-based habits which don't apply to the new systems anymore.
So you need to know the answer to the following question:
What are the Linux alternatives for the most common maintenance tools and files on NetWare?
Instead of typing load monitor, you use the Novell Remote Manager (NRM).
Probably you're already familiar with NRM, which is also available for NetWare.
Use a web browser to access http://server_name:8008 and you're in NRM.
It almost explains itself while you're browsing around there, of course it's also thoroughly documented on the Novell website so I don't bother to discuss it any deeper at this place.
For more information on NRM, check the OES documentation, section management tools, 'Novell Remote Manager Administration Guide'.
Command-wise it's just one character more: use the command ndsrepair instead of dsrepair. NDS, isn't that very, very old? Yes indeed, NDS goes a long way back and since ages we've become used to the name of eDirectory. Somewhere on the Novell site I read that eDirectory is the name for the product and that NDS is still used for the technology.
Anyway, however you call it, you still need to know what the dsrepair.nlm alternative is on Linux. For the options with ndsrepair you can simply browse the manual pages by typing man ndsrepair.
If you're just planning to perform a regular NDS health check, you better use this Novell TID: 10060600 (I secretly suspect that every Novell admin knows this number by heart). It describes a complete step-by-step instruction for performing an NDS health check on any platform you'd like.
On Linux you use the VI text editor. One way to do it is: just go to file path and type vi with the filename behind it.
The file will be opened in view (read-only) mode. If you want to change anything, set the cursor on the place where you'd like to change something and use INS to get into write-mode. Press ESC if you want to go back to the view-only mode.
If you'd like to search anything in vi, type a forward slash with the search entry and hit Enter.
To quit without saving, type :q! and press enter.
To save and quit, type :wq and press enter.
These are just some common tips for using vi. Type vi -h to get more information about using vi.
standard server management and monitoring
For a lot of monitoring tasks, such as server health or disk space monitoring, you can use NRM which I already discussed some paragraphs up above.
If you'd like to monitor your systems in a more detailed way, with more functionality and centralized (so you can monitor all the servers and - more important - their services such as e-mail functionality, printing et cetera), consider using a specialized monitoring tool.
If it can't cost a dime you can use Nagios which is free and which will monitor almost everything possible.
user management and other NDS related management
For user management, eDirectory management and such, you use Novell iManager.
You should also be already familiar with iManager.
Use a web browser to access https://imanager/nps.
For more information on iManager, there is very outspread documentation on the Novell site. http://www.novell.com/documentation/imanager26/
For management and configuration of GroupWise agents, users, link configuration and so on, you can still use good old ConsoleOne.
- GroupWise Agent screens on the server
Concerning the nice blue agent screens (I mean, blue by default. But I don't know anyone who ever made his GWIA agent screen look brightly pink so I just assume that everyone who reads this never did mind about changing screen colours) on your NetWare server: you've lost these blue screens.
I know, you're going to miss the sweet little bracket in the left-top corner turn around and around, giving you a warm, safe feeling of assurance that the agent still was running. But in a while you'll get over this and you will realize that it's quite an understandable decision of the Novell guys to dump these screens which you always tried to give a nice descriptive name but which you always ended up calling 'blue screen' because you couldn't think up another name, no matter how hard you tried.
Instead of the agent screens you'll use the agent web interfaces.
As you're a GroupWise Administrator, you should already know how to access these interfaces. By default you can browse to http://agent_server_ip_address:9850 for GWIA, http://agent_server_ip_address:7180 for the first MTA and http://agent_server_ip_address:7181 for the first POA.
You can find the right port numbers in ConsoleOne, properties of the agent, GroupWise tab, Network Address tab.Besides that you can install and use GroupWise Monitor for further monitoring tasks on the GroupWise services.
- Other ways to manage GroupWise
You can also use the command line for a quick check on the GroupWise system.
Go to the /etc/init.d directory where the grpwise daemon is installed by default.
Type ./grpwise status and the status of all the agents will be displayed.
To stop an agent, type ./grpwise stop with the agent name behind it.
Don't forget to put the agent name behind it, or all the GroupWise agents will be stopped!
In the same way you can start an agent by using the ./grpwise start command with the agent name.
You can restart all the agents by typing ./grpwise restart
This is just a quick and very basic explanation how to use the command line for basic manipulation of the GroupWise agents.Of course the agent log files are also still a most valuable source of knowing what happens in your GroupWise system. Check them regularly.
Cluster specific commands
If you've worked with cluster services on NetWare, there shouldn't be any surprises for you when working on the Linux command line.
You'll already know a lot of the commands you can use.
Type cluster status to see the states of the cluster resources and which nodes they're running on.
You can also use watch cluster status to see the status continually, the list is being updated. Use CTRL+C to stop this.
Type cluster help for all the possible commands and switches.
abend.log, sys$log.err, health.log, console.log, and so on
No more bad dreams about SYS:/SYSTEM/ABEND.LOG for you! In fact, since Linux, the ABEND.LOG is history. It has ceased to be, it is an ex-abend.log, it is... pining for the fields. Pining for the fields? Yes, because now you have a very neat set of new log files to scan!
In Linux, most of the log files you need are to be found in /var/log.
A very useful log file you can find there is messages. Just open it with vi and check what's in there!
One level higher, in the /var directory, you can also find specific log files in their individual subdirectories. For example most nds-related log files you'll find in /var/nds.
There are some nice script based utilities to gather log data from your server. You can find these in other cool solutions, articles and tips on the Novell web site.
Load and unload
On Linux you don't load a module, but you run a script which probably runs other scripts.
The most services are run from the directory /etc/init.d.
In most cases it's possible to start the service by using a command like ./service_name start (if you're in the path) and to stop it using ./service_name stop
This depends on the service.
An extra tip
You might want to take a look at the OES documentation, section 'Linux Tips for NetWare Administrators'.It maps all the NetWare commands to their Linux alternates, which I find very useful.
This article is just some basic information gathered from different sources.
I put the information together so NetWare admins who are relatively new to Linux have a quick reference for the most basic tasks.
I strongly recommend that everyone who administers a Linux environment should follow relevant courses on the subject.
Never perform any action on a server when you don't exactly understand what you're doing or why you're doing it!
Novell Cool Solutions (corporate web communities) are produced by WebWise Solutions. www.webwiseone.com