Hello,
I'm a Solaris Admin by trade and we are getting more Linux server in. I was wondering what command Red Hat uses for hardware faults?
lets say you see the fault light on the server what command do you use to see what the fault is?
for example the Solaris command is fmadm faulty. this commands gives info on fan speeds and other low level faults.
Any help would be great.
Depends on the hardware. HP, for example, has a suite of command line utilities that interface with the iLO and can query the hardware status. Likewise for Dell servers.
it is a Sun x86 Blade. I tried the Solaris command but it will not work in Red Hat.
Thanks for your help
As was pointed out, this is a problem for your hardware vendor. This is not a Linux issue.
Everyone,
Well I did some checking and it seams that RHEL doesn't have a built in feature that will check for hardware faults like Solaris does. I was just trying to get better with my RHEL skills. I already knew how to get to the fault information from the lights out manager.
---------- Post updated at 04:42 PM ---------- Previous update was at 04:18 PM ----------
I have found the answer. I posted the same question here this is the answer I got.
Red Hat or CentOS CMD - Linux / UNIX Forum
The command is dmesg
Ok, this might come off a bit harsh, and I want to apologize in advance: Did you even really read the above answers, or did you just skim them without any attempt at making sense of them??? Of course the Solaris command won't work in Linux, because it's a Solaris command. There is no general command to check for hardware failures in Linux, because that's the hardware vendors job!! And the dmesg command won't give you any information on fan status and whatnot, as it can only report the errors the kernel sees. And without kernel modules provided by the hardware vendor it won't even know it has a fan that could fail!
Disclaimer: It has been approximately 7 years since I used Linux regularly. However, I would be surprised if any of what follows is no longer applicable.
You should be aware that dmesg simply prints out the contents of a kernel log buffer. It is a ring buffer and it's not very large. New messages overwrite older ones once the buffer has been filled. If your kernel is feeling chatty, you could very well miss messages if you depend on dmesg.
dmesg can be convenient to check the latest kernel messages, but I would not consider it a comprehensive reporting tool in and of itself. For that you need to look elsewhere.
For example, soon after booting a linux machine, dmesg will print out all the startup messages spat out by the kernel during its initialization and hardware detection routines. Try dmesg a day later (or a minute later if the kernel has encountered some issue) and it's possible that the boot messages are now gone (they should still be available in a log file, even if they've been overwritten in the kernel buffer read by dmesg).
Regards,
Alister
alister thanks for your informative answer.
dmesg seems not to be what I was hoping it was. I guess RHEL just doesn't have this functionality. I have been doing a lot more Red Hat stuff lately and a lot of stuff seemed easier to do then on the Solaris server I normally work on. I just thought if Solaris can do it surly RHEL can do it too. I guess not.
Anyway I never asked how to get the faults or errors from the hardware I am using, only can RHEL access the info and if so how. The RHEL server is on a Sun Blade so there are two ways to get to the info. One way is thought the Lights out manager and the other is console port.
fpmurphy I looked into the lm-sensors program but I don't like installing 3rd party stuff and prefer native OS tools if possible. Thanks for helping me though.
Anyway thanks for everyones input.
lm-sensors should be available from the Red Hat repositories, and is as much 3rd party stuff as bash, Perl, Apache, or NFS. Each of these tools is made by different people, and at most 1% of those is employed by Red Hat. Not even the kernel is made by Red Hat, they're contributing to it, but the main work they do is package all this software together and glue it together.
Yes I know lm_sensors is in the repositories I checked. I also saw that I have some files to configure to get it to work. That would take more work then I am willing to put into it at this time. I will just use the Lights out Manager for now.
I understand that most of the Linux stuff is made by other people other then Red Hat. If the package/software is not part of the standard build I have to get special permission from security to install it. I'm mainly a UNIX admin and I don't want to add any programs that may not be needed and that I will have to secure later.
Thanks for your help. Maybe I can be of help to you guys at a later time.
The reason that Solaris can give you that information is that the hardware and the software are being provided by the same vendor. Linux runs on commodity equipment, which has many different parts. It would be foolish to assume that the same integration exists, but using the tools from your manufacturer will allow you to do the same things, as they should provide you with the integration tools to check hardware, etc.