Dude, I'm getting a dell....if CRUX will support it that is....
We are considering the purchase of a DELL 2950 Power Edge server. Does anyone have any experience getting the PERC5/i RAID controller working with Crux2.3? Ori
Ori Johnson wrote:
We are considering the purchase of a DELL 2950 Power Edge server.
Does anyone have any experience getting the PERC5/i RAID controller working with Crux2.3?
Ori
Hardware support has little to do with the distribution you are using. If the linux kernel supports that controller, crux does. I don't know with what options the kernel on the install CD has been compiled with, but you might need to remaster the install CD. Glenn
We are considering the purchase of a DELL 2950 Power Edge server.
Does anyone have any experience getting the PERC5/i RAID controller working with Crux2.3?
For this controller you'll have to enable CONFIG_MEGARAID_SAS in the kernel. There seems to be a DMA bug with firmware versions less than *-0095 [1], so if you happen to get one of these, try to update the firmware version, or follow this workaround [2]. [1] http://www.ussg.iu.edu/hypermail/linux/kernel/0609.0/1387.html [2] http://marc.info/?l=linux-poweredge&m=115984680103611 Regards, -- Alan Mizrahi
The kernel can see the drives, and I am able work with /dev/sda and /dev/sdb (the logical drives under the RAID controller) right of the Crux 2.3 CD :) Here is the problem that I currently have: I want to monitor the health of the RAID arrays. After booting the system, I removed one of the drives from the Array. I was expecting to see an entry in /var/log/messages. The RAID controller handled the loss of a physical drive correctly (the logical drive was still accessible), but no messages were logged in /var/log/*. I then rebooted. When the system initiated the RAID controller, the controller stated that the logical drive has been degraded. This is what I expected to see...but that same message does not appear in dmesg or /var/log*. I then replaced the drive and the RAID control re-synchronized the drives correctly. So, my question is: How can I get messages from the RAID controller that there is a drive failure in the RAID system? Thanks, Ori -----Original Message----- From: crux-bounces@lists.crux.nu [mailto:crux-bounces@lists.crux.nu] On Behalf Of Alan Mizrahi Sent: Thursday, November 15, 2007 4:31 PM To: crux@lists.crux.nu Subject: Re: Dude, I'm getting a dell....if CRUX will support it that is....
We are considering the purchase of a DELL 2950 Power Edge server.
Does anyone have any experience getting the PERC5/i RAID controller working with Crux2.3?
For this controller you'll have to enable CONFIG_MEGARAID_SAS in the kernel. There seems to be a DMA bug with firmware versions less than *-0095 [1], so if you happen to get one of these, try to update the firmware version, or follow this workaround [2]. [1] http://www.ussg.iu.edu/hypermail/linux/kernel/0609.0/1387.html [2] http://marc.info/?l=linux-poweredge&m=115984680103611 Regards, -- Alan Mizrahi
Ori Johnson schrieb:
The kernel can see the drives, and I am able work with /dev/sda and /dev/sdb (the logical drives under the RAID controller) right of the Crux 2.3 CD :)
Great...
Here is the problem that I currently have: I want to monitor the health of the RAID arrays. After booting the system, I removed one of the drives from the Array. I was expecting to see an entry in /var/log/messages.
I don't know your driver/raid combination, but if you know that you usually get more log output, try to configure your kernel / system log daemon to dump out everything it gets: man syslogd man klogd For kernel development, I also add a line like: *.* /var/log/all to /etc/syslogd.conf and restart syslogd to get everything to /var/log/all. If the driver or kernel doesn't emit enough debugging information, it's propably a good idea to configure your kernel (the latest one) according to your needs (turn on more debugging). However, customizing your own kernel is propably not trivial if you are not used to it... Regards, -- Clemens Koller __________________________________ R&D Imaging Devices Anagramm GmbH Rupert-Mayer-Straße 45/1 Linhof Werksgelände D-81379 München Tel.089-741518-50 Fax 089-741518-19 http://www.anagramm-technology.com
Hi, I'm using a software raid with mdadm which has monitoring and reporting built in. I'm not sure if you can use it to watch hardware raid setups. Joe
Joe Gilmour schrieb:
Hi, I'm using a software raid with mdadm which has monitoring and reporting built in. I'm not sure if you can use it to watch hardware raid setups.
Ah, right... we were talking about a HW raid here - my bad. That should be transparent to the kernel, so no log output, no mdadm here. Nevertheless, there should be a userspace tool for the hw raids at least from the raid manufacturer. A quick google search led me to this brief overview: http://wiki.debian.org/LinuxRaidForAdmins You are then looking for the cmd line tools for your controller. Regards, -- Clemens Koller __________________________________ R&D Imaging Devices Anagramm GmbH Rupert-Mayer-Straße 45/1 Linhof Werksgelände D-81379 München Tel.089-741518-50 Fax 089-741518-19 http://www.anagramm-technology.com
participants (5)
-
Alan Mizrahi -
Clemens Koller -
Joe Gilmour -
Ori Johnson -
RedShift