Checking RAID status on Dell / MegaRAID
I just got a Dell Poweredge 1800 server and I'm trying to run Slackware 10.1 on it. The server has a PERC4/DC SCSI 320 RAID controller, with 2 drives in a RAID 1 configuration.
Slackware install went great. Using the raid.s kernel, everything installed just right out of the box. Seems to use the megaraid driver, and the PERC4/DC controller gets recognised as an LSI Logic MegaRAID LSI53C1030. I built myself a 2.6.11.3 kernel and everything still seems good. My only question is: How do I monitor the RAID status? I would like to know how to tell if a drive has failed and needs replacement. Is there a utility for this, that I can run on Slackware? Dell shipped the machine with Red Hat Advanced Server (in the box, not pre-loaded) but I wanted to see if I could get it running on Slackware so that it would be consistent with our other Linux boxes. I couldn't find anything for Slackware on Dell's support site, since they don't officially support Slackware. There was one utility that looked like a RAID monitoring thing, but it was only available as an RPM for Red Hat. So, any monitoring tools anyone knows about? Doesn't have to be fancy or GUI or anything. Just enough to get the job done. |
I don't use hardware raid myself, but I believe there are some pseudo-files in /proc for monitoring a raid array.
|
I was looking into /proc for something like that. I know that ATA RAID cards I've used in other machines report status in a file in /proc/scsi.
But I didn't find anything for this card. My /proc/scsi has the following: -r--r--r-- 1 root root 0 2005-03-15 04:50 device_info dr-xr-xr-x 2 root root 0 2005-03-15 04:50 mptscsih/ -r--r--r-- 1 root root 0 2005-03-15 04:50 scsi and those files contain "capabilities" type info, nothing I could see that was like RAID status. There is also a /proc/mpt which contains: mpt: dr-xr-xr-x 2 root root 0 2005-03-15 04:55 ioc0/ -r--r--r-- 1 root root 0 2005-03-15 04:55 summary -r--r--r-- 1 root root 0 2005-03-15 04:55 version mpt/ioc0: -r--r--r-- 1 root root 0 2005-03-15 04:57 info -r--r--r-- 1 root root 0 2005-03-15 04:57 summary Again these seem to be "capabilites" and "model number" type of info. The "info" file contains: ioc0: (f/w download boot flag set) ProductID = 0x0202 (LSI53C1030) FWVersion = 0x01032300 (fw_size=40048) MsgVersion = 0x0102 FirstWhoInit = 0x00 EventState = 0x01 CurrentHostMfaHighAddr = 0x00000000 CurrentSenseBufferHighAddr = 0x00000000 MaxChainDepth = 0x30 frames MinBlockSize = 0x20 bytes RequestFrames @ 0xf7dc2000 (Dma @ 0x37dc2000) {CurReqSz=96} x {CurReqDepth=255} = 24480 bytes ^= 0x7000 {MaxReqSz=96} {MaxReqDepth=255} Frames @ 0xf7dc0000 (Dma @ 0x37dc0000) {CurRepSz=64} x {CurRepDepth=128} = 8192 bytes ^= 0x2080 {MaxRepSz=128} {MaxRepDepth=255} MaxDevices = 16 MaxBuses = 1 PortNumber = 1 (of 1) And the "summary" file simply says: ioc0: LSI53C1030, FwRev=01032300h, Ports=1, MaxQ=255, IRQ=66 Any ideas where else under /proc or /sys that I might find some RAID status type stuff? |
Did you try rpm2tgz; move result to suitable tmp directory and then explodepkg on that result file. Dell's own package could meet your Slack needs. Chances aren't great it will work on a non redhat patched kernel but might give you clues to fiinding your answer for slackware.
Just a thought |
I think I found the answer. LSI Logic has a command-line utility on their website for monitoring and rebuilding the RAID. They really don't make it obvious on Dell's or LSI's web sites that this is what you need; I found out about it digging through message board archives.
Anyway the utility is called megarc. It's only supplied as a binary, but it executes just fine for me under Slack 10.1 on the Dell PE1800 box. It gives text output, and I can run it from a cron job and grep the output for certain strings, and send out e-mails for failed drive notifications. Here's a link to the megarc utility http://www.lsilogic.com/downloads/do...nload_type=all |
Hi MadRabbit,
I have a similar configuration with yours. My server is Dell PowerEdge 1800, dual XEON 3.00GHz, 2GB memory and the Dell/Adapter RAID SATA 6 channels card with 6 SATA 150 HDs. Since the server have just been purchased and I am new to the RAID configuration and monitoring, I am trying to test for RAID 1. The configuration of RAID 1 is OK. RAID 5 is my target approach. The DRAC Card was not purchased. The target server OS is Redhat Linux Enterprise AS 4. However, the supported OS verion for Redhat Linux Enterprise is only up to 3, with reference to the enclosed setup CD's documentation. Now, I have two problems: 1. How to monitor the server health status and perform administrative tasks remotely without installing DRAC Card? As far as I know, the Server Administrator can do the task. However, I could not access the server with using a browser on port 1311. The OS is the Redhat Linux Enterprise 3. It had response and asked me to download an unknown file. The file name was different for each time when I tried to access. I guess the server port 1311 was being used by the Server Administrator. 2. When I install Redhat Linux Enterprise AS 4, the ethernet card could not be recongized. Even I downloaded and installed the driver that was downloaded from Dell's support site with no dependency, the ethernet card was displayed in the Hardware Browser, but could not be configured. I would be appreciate if you or anyone give me a helpful suggestion. Thanks a lot! Bruce |
I'm not using Red Hat so I'm not so familiar with the remote administration on it. We are running Slackware on both our Dell boxes.
If the RAID monitoring package from Dell won't work on your version of Red Hat, I suppose you could always fall back to the megarc utility like I did. It's command-line based so you can use it remotely via ssh. |
anyone know where to find megarc.. lsi removed the link on their website
|
|
I've had two different LSI RAID cards in my Slack server. You can use all of the following against the cards:
MegaMGR 3.8 megarc 1.11 megamgr 5.20 They all perform slightly different functions. megamgr is a GUI, megarc is command line, and MegaMGR is a daemon that monitors the RAID. All of these use a script to drive the binary portion of the code. That script needs modifying a little, as the RAID cards no longer populate a major device node, they use a minor. I can post diffs for all 3, if required. Also, all 3 of the binary portions are statically linked, so they work on both the 32-bit, and the "pure" 64-bit versions of Slack. ***** Update ***** I just spotted, you are on Slack 10. The scripts work fine on that version, without modification. BTW, when you start the GUI, don't assume it's frozen, if it takes a while before any options appear. That's quite normal. ***** 2nd Update ***** Jeez, I just noticed that this thread was resurrected from the dead. C'mon folks, what are you thinking. Cheers. |
All times are GMT -5. The time now is 08:17 PM. |