SAN storage systems fail rarely and at a high cost: these systems have many shelves, fibre channel fabrics, LUNs that are created and then assigned to specific hosts that can't find them. EDR views all SAN work as structured engagements that are scoped in from the start. We have an NDA with every client as standard procedure. Then we use disk images to engineer our solutions, because once things get this big, improvising is not affordable.
The SAN has complexity through levels of abstraction (indirection): physical disks are clustered together into pools, those pools are then divided into LUNs, which are then mapped out to hosts using Fibre Channel or iSCSI protocols for the hosts to use as local disk drives, and frequently there's another level of virtualisation on top. If shelves fail, if the controllers disagree, or if an administrator makes some changes that orphan a LUN, the actual data will likely be fine; however, the mapping to where this data exists is broken. The recovery process typically starts by imaging all members involved in the system, then creating a topological map by identifying all the pools and LUN geometries via on-disk metadata, and finally rebuilding the host file systems or the VMFS/CSV volumes that were expected by the hosts. For multi-TB size estates this occurs in phases with regular checkpointing points at which you formally agree to move forward.
SAN jobs are initially a conversation, not a courier: the make and model of arrays and shelves, number of disks on those shelves, what had been changed prior to the failure and which was the highest priority data set. From there comes a detailed scope and a fixed price for the project, typically larger in scope and higher cost than a single-drive job, as everything will be broken down into line items to help you understand why. You then remove all disks from their respective shelves (each labelled per shelf and slot) and ship them insured or drop them off at our facility, Monday to Friday, 9am to 5:30pm, with no appointment needed. Discretion is standard operating procedure: all clients have NDAs signed, client's data is treated exclusively by our engineers, and communication with clients regarding status will always go directly to a designated contact. We will also work cooperatively with a client's existing IT support team or storage vendor if required; again, our focus remains solely on the recovery of the data and not on internal politics.
This is a typical, standard SAN-recovery brief. A vendor will restore an array’s functionality but it can’t provide the history of the array. The recovery process creates a copy of each disk at a point in time (the image) that allows for reconstruction of LUNs into their prior states, targeting the data your backup misses. Therefore, you should recover with both: partial backups along with recovery beat either alone.
Yes, staged: members are imaged in priority order, geometry is recreated and the most critical LUNs are restored first, which allows for access to your highest-priority data sooner while the rest of the long tail processes. While scaling may adjust both the timeline and the price tag associated with the restoration effort, it does not affect the methodical approach involved.
Routine: a large share of SAN work revolves around virtualisation storage. As soon as the LUN has been restored, VMFS is restored and each VMDK (virtual machine) can then be individually recovered and tested. The virtual machine page will cover everything happening inside of them.
First, a free diagnostic, then a fixed written quote before any work is done. This means no surprises on the invoice (because we don't give you an invoice until you approve the number).
While SAN-specific write-ups are still rare in our case files, the same principles apply as were used to save this RAID 0 server. The principle being: image each member prior to attempting to rebuild anything live.
It will take only two moves to place your device at our service area, and once that has been accomplished, the clock on your free look starts.
First get a quote, over the telephone or using the online quotation form. When sending your internal drive, remove all cables, caddies and power supplies, as they will have little use when sent to us. But when sending an external drive, send the complete item, including its original casing, together with the supplied cable and power supply. Ensure that the device is securely wrapped and unable to move during transportation. Either post it insured to 4 Redheughs Rigg Westpoint, South Gyle, Edinburgh EH12 9DQ or call into our offices Monday to Friday, 9am to 5:30pm, with no appointment needed. Please include a brief description of how your device was damaged. If you have printed a shipping form, please attach it along with the device.
Once the arrival of your device has been confirmed, the diagnostic will run free of charge, and then when we make the phone call to you, we'll name the problem and tell you the fixed price to fix it. The bench will remain dormant on your device until you give us permission to proceed.
A free diagnostic, a written price quote by array, and all disks imaged read-only prior to beginning reconstruction.