As with everything else IT, you will need to maintain your SAN and connected equipment, to ensure optimal performance, availability, and functionality. Here are just a few of the procedures to be performed and a quick look at some preventative measures to be considered.
Sidebar: My original intent with this series of posts was to only provide gumi center gaberje a basic recipe of setting up an uncomplicated SAN, but it has grown into providing additional information and procedures as well. Throughout this series I am only providing the basic information only. You still will have to get hold of the technical and theory of operation documentation, and then study those until you have a clear understanding of the principles and practical application before you will be able to build, maintain, and optimize your SAN with confidence.
Keep in mind: The device software is continually improved, gumi center gaberje and you only should use the documentation specific to the software version currently installed on your equipment. About events, logs, and files on the V7000 and CISCO switches
Our SAN has been installed, gumi center gaberje commissioned, configured, as per the previous posts, gumi center gaberje and now is running in full production mode. Life is good. But we all know that sooner or later something will go wrong. What do we do then?
If your kit is still under warrantee, or is on a maintenance contract, you log a call to have it fixed. But the very first thing the maintenance gumi center gaberje provider request is for you to upload the device logs to them.
Both gumi center gaberje the V7000 and the CISCO switches will log events which indicate abnormal processing. Information about these events is stored in event logs or error logs, or whatever the manufacturer chooses to call these logs.
At the same time you also would like to have available a copy of the setup information as well as the configuration of the equipment, so that you may load it back after the broken unit has been fixed or replaced. Updated diagram of the SAN
Looking at the diagram of our SAN, to which I now have added Ethernet Switch A and Ethernet Switch B, you will see that we have an Ethernet connection from the switch (the purple arrows) to every CISCO SAN switch, the V7000, and a server at each of the Data Centers.
These Ethernet switches must be isolated so as to form a stand-alone LAN, and thus must NOT be connected to the Data Center network! This done for security purposes, as your server ILOM/ALOM/ELOM management ports and SAN equipment management ports are all connected to this private gumi center gaberje LAN, and you definitely do not want to expose these ports to the general network and so allow access for hackers!
For you to reach this private gumi center gaberje LAN, so as to allow you to remotely manage the server and SAN kit, connect a spare Ethernet network port (NIC) on two of your servers to your private LAN, and allow no automatic routing to this network. You thus have to remotely log onto your server, and only then can use ssh to connect to the SAN kit or server ILOM.
It is a very good idea to regularly do a manual check of your event/error logs, even if you did set up automatic e-mailing of events, as e-mail is not a failsafe mechanism for delivering gumi center gaberje messages. Manually checking an IBM V7000 Storwize
Chapter 9 of the Clustered system gumi center gaberje diagnostic and service-aid Commands chapter of the SVC and V7000 command line interface users guide manual lists all the available event log commands. Please read that chapter carefully to get an understanding of the log functions available. Manually checking a CISCO MDS 9148 SAN switch
But definitely do have a look at Chapter 52. Configuring System Message Logging of the Cisco MDS 9000 Family CLI Configuration Guide for more information regarding checking the status of your CISCO MDS 9148 switch, as there are multiple options and methods available. Manually checking a SUN/ORACLE Solaris gumi center gaberje server
In this section I will be focusing on the logs as requested by the manufacturer/service agent. It is up to you to read the user guides to get a more detailed understanding of what logs can be generated. Just keep in mind that not all error logging is enabled by default some logging you have to explicitly enable. But also know that if you enable all error logging possible, you will have to manage the logs carefully as the available memory, where these logs are stored, is limited.
Oh! And remember to check and set the time and timezone on your SAN kit. If your system time on the V7000 or CISCO SAN switches is not correct, total confusion can ensue due to events shown in the error/event logs that do not coincide with what happened on other equipment! IBM V7000 Storwize configuration file
Please note: Some commands are used when you have a SAN Volume Controller (SVC), i.e. lsnodehw for SVC, and lsnodecanisterhw for non-SVC. Alternate commands can be found in most cases for when you are running the V7000 without a SVC. Have a look at the documentation listed above and find the cor
No comments:
Post a Comment