Xserve Raid not appearing in Raid Admin

You can ping the network then guess. You can check your DHCP server and determine the address by matching the MAC address of the controller. You can reset the controller on the XRaid (do that after unmounting the volume). You can try to simply Add a new Raid to Raid Admin (as your address may have changed if it is configured via DHCP).

The red drive is probably a failed drive. If you have raid 5 with spare, then the spare should be rebuilding. If you have raid 5 and no spare, then order a replacement and replace the dead drive.

Hope this helps.

Ping your network to talk to all devices on the subnet. For example, if you have a network of 192.168.0.0/24 (192.168.0.1 - 192.168.0.154 usable) then you can ping the broadcast address of 192.168.0.255. All devices, unless defined to not respond to broadcast ping will return a response. As for the MAC address, you need a flashlight, good eyes, and the ability to see the rear of the unit. The MAC address is printed on labels on the rear of the case behind the power supplies. With the MAC address, you can reference your DHCP server and find the address. If the system has static addresses, then you will not see it listed in your dhcp server.

Hope this helps.

Ok, so let's start with the basics.

Ethernet is needed only to manage the Raid.
Fibre channel is how the volume is mounted on the host.
Lack of connection via ethernet will not prevent volume mounting.
RAID is only a hardware redundancy and problems can still occur.
There are many levels of RAID, each with its own level of redundancy (from none to high)
You do have a backup, correct?

So, why can the server host not see the management connection (ethernet) of the RAID? There are a number of possible causes. Your switch might have a VLAN configured. Your Raid may be on a different subnet than the host (What is the IP address of the host? What is the IP address of the RAID?). Regardless of the cause, the fact that you can connect using your laptop means that you can manage the unit and identify the problems. So, do this. Connect to the RAID for management (over ethernet) and determine the type of RAID that you are looking at. Is it a RAID 1, 2, or 5?

Next, go back to your host system. Open System Profiler. Click on the Fibre Channel section. Can you see the RAID? If yes, then open Disk Utility. Can you see the physical device and the volume? If the volume is not mounted, can you mount it by clicking the mount button. Do you get an error? If so, can you repair the volume using Disk Utility?

Hope his helps

Ok, just confirming. The left channel is a RAID 5. The right channel is a RAID 5. Then you are using Disk Utility to mirror the data across the channels. This is a software mirror. If this is the case, then you "should not" have lost your data. First, the RAID 5 provides one of the highest levels of drive level redundancy. A RAID 5 with no spare can suffer the loss of a single drive and still be valid. Based on your description, you have this on both sides of the XRaid. This means that you can drop a drive on each side of the case concurrently and still be ok. But, on top of that, you are saying that you have a software mirror. If this is the case, then the data on the left side is the same as the data on the right side. If either underlying raid 5 was to die or go offline, the mirror should degrade but data access should be fine.

The way you are describing this, I fear that the software raid is a stripe, not a mirror. Do this to determine:

1: Open Terminal
2: Type:
diskutil checkRaid
3: What is the result you get back?

Hope this helps

Ok, the prognosis is getting grimmer. However, there is a chance that your data is still there. Recall, you have RAID 5 as the underlying format and in theory, the loss of one drive in a RAID 5 will not destroy the volume. Try this:

1: Shut down the Xserve
2: Make sure both fibre channel cables are plugged into the Xserve
3: Using the Raid Admin tool, connect to the XRAID and reboot the unit, effectively reseting both controllers.
4: Wait for the unit to complete it reboot process (you may get an audible alarm since you have a failed drive)
5: Start the Xserve
6: Cross your fingers and maybe the volume will come back up.

Hope this helps. The use of a stripe is a fearful prospect. I understand for certain deployment scenarios like video ingest etc. However, it looks like someone was trying to make a single large volume and used the stripe as a solution.

Hope this helps

adam taylor1 wrote:
As the data is backed up, I am not worried about having to re-format the entire Raid. When (and if) I do - what is the best way to configure the set-up for video work and retain the Raid 5?

That is a loaded question 🙂 It all depends on a nearly infinite number of details. What type of video are you shooting in? DV, HD compresses or uncompressed, etc. How is the storage used in your workflow? Ingest, editing, play out. How many people are connected to it and how do they connect? Is this direct connect? Is it a reshare over AFP? Are only project files kept there and all ingest/cutting is done on the workstation? And those are just the tip of the iceberg.

Here are some basic suggestions. RAID 5 is the best format for relative safety of your data. However, the nature of the XRaid design is that you end up with two independent volumes. So say you have 14 500 GB drives. This is 7 TB of raw storage. Doing a Raid 5 on each side, no spare, will yield 3 TB (about 2.7 TB formated). On top of those formats, the unit can be software raided into a raid 50 (two raid 5s striped) or 51 (two raid 5s mirrored). These are tenuous as you've discovered. The later is wasteful as you throw 7 TB of raw storage at a volume that yield 2.7 TB of usable space. In high availability/redundancy situations this makes sense. But the amount of data that needs that level of protection is probably pretty small. The other option, which is what I guess happened to you, is taking the two raid 5s and striping them together. The only benefit here is that you get what appears to be a single volume spanned across the two controllers. There is a perceived speed benefit here and since you mentioned video workflow, that may explain the decision. However, unless you are doing ingest, this formating likely makes no sense.

Hope this helps