Most enterprise storage outages happen as part of a series of events – i.e., a failure of one disk leads to another disk's rebuild, which causes a failover to occur but the failover doesn't fully complete. Reconstruction works from the bottom up through every layer a vendor has stacked, so each requires the previous layer to be correct. This is why we begin by having a conversation about your needs rather than providing a quote.
Configuration details such as LUN layouts, host mapping information and pool metadata may be just as important as what disks were involved when trying to recover your data. Export the configuration or take pictures of all of the management pages while you still can, and do not let the system rebuild, reinitialise or expand while a fault is unresolved.
Storage area network (SAN) recovery quotes are provided after a discussion with a customer rather than by referencing pricing information. Most discussions regarding SAN recovery are brief if you have five pieces of relevant information available.
Dell EMC, HPE, NetApp, IBM, Hitachi — and which family. Each generation of hardware products from these vendors differs sufficiently in design configurations to impact estimates.
Details about group sizes, RAID level, number of shelves and if there were hot spares being used and if they had begun rebuilding.
LUNs existing, rough size of them, and the host(s) they were mapped to. An export from the config tool would be great, but a photo of the actual screen can also help us.
This will greatly impact our reconstruction as we now have to track both physical block allocation and virtual/logical address space as well.
A rebuild, an expansion, a failed controller fail-over that didn't finish, a reinitialisation...this is the single most important piece of information you can give us & the hardest to get out of you.
SAN failures usually happen in sequence as opposed to just one event. That is also why the account of what actually occurred matters.
Beginning with the lowest level and moving upward by layers, no steps can be missed; therefore it is quoted as a project.
Every disk (across all shelves) was first imaged read-only, which can take several days for large arrays. Then from these read-only images were constructed RAID groups using derived geometry from the data itself instead of depending upon vendor provided metadata describing a condition that likely does not exist now.
At the next higher level, the storage pool will be mapped. Then, the boundaries for each logical unit number (LUN) will be found. Following this, the thin-provisioning metadata will be used to determine the actual location of allocated blocks. Finally, after determining how the allocated blocks fit into the storage pool, either a file system or a VMFS datastore becomes visible. Therefore, because each subsequent layer depends upon its previous layer having been properly created, errors made at the bottom-most levels result in errors appearing far up the chain as nonsensical data — which is why reconstructions are verified at each level before proceeding to the next instead of waiting until completion.
The cost for SAN or enterprise storage recovery is priced separate to the cost for server or RAID work because the job is much larger.
Enterprise storage and SAN recovery can be quoted from £1,250 +VAT. We do not quote off a price list. Every item we consider will affect your final quotation; i.e., number of members in the storage unit, number of shelves in the storage unit, vendor layout, what has already been attempted, etc. We also require a 50% deposit when services are provided. The final payment is made upon successful completion of our services.
We operate under confidentiality by default. NDAs are signed as a matter of course. One person will act as your main contact for your project. Data remains within the United Kingdom and working copies are destroyed in accordance with your designated schedule.
You send us the disks associated with the failed pool and inform us of the array make/model and RAID configuration. There is no cost for the initial diagnostic process. Once the diagnostic is done, we give you one per-array figure in writing prior to beginning any work.
All diagnostics begin once we have received the disks from the pool — labelled in slot order — along with information about the array manufacturer and type. Please pack carefully and include your contact details. There is no charge for the diagnostic. One price will be given in writing prior to commencing any work.
Posting it? Use a tracked and insured service. The value of what is on the drive far exceeds the cost of postage. Bringing it in? Weekdays from 9am to 5:30pm, and it still needs packing as described above for the journey.
Do you need a scope of work done before sending it? Do you want to know how much it will cost first? Fill in the form and describe your RAID setup (including layout), and what you think is wrong with it. An engineer will then review your submission and reply back to you with a custom quote based on what they find.
We’ll be in touch shortly. If it’s urgent, call 0161 871 0788.
The most common SAN recovery enquiries we receive.
More layers with each depending on the one below. Each layer must also be rebuilt. The layers include: disks in RAID groups; RAID groups as part of larger pools; the pools are then divided up into thinly provisioned LUNs assigned to hosts. A file system cannot be seen until all of the layers have been rebuilt.
Typically. This pattern usually indicates that a pool or LUN metadata has been lost while the actual disk storage remains. This means that the data exists and the way the storage space is being defined is missing. It will need to be reconstructed based upon images of the individual members.
Your vendor, model number, RAID configuration, LUN mapping details, whether thin provisioning had been activated, and what has already been run. Of these listed items, what has already been run is by far the most important and would significantly reduce the time required to complete the job.
Yes — the VMFS file systems that contain those VMDKs and the individual VMDK files themselves. The actual reconstruction of the LUN itself will also need to occur before the datastore can be restored. Hyper-V volumes are treated similarly.
Yes, by default. We name a single point of contact for each client and keep all their data within the UK. Enterprise clients usually insist upon signing a non-disclosure agreement so regardless of whether they ask for it or not, we do.
From £1,250 plus VAT — quoted after a scoping consultation. This amount is based on the number of members in the SAN, how many shelves are configured and what type of format the SAN uses. A 50% deposit is required with the remaining balance due on successful recovery.
Quotes are done per unit as the number of members in an enterprise environment, how many shelves are connected together to create one environment, and how the hardware is laid out by each vendor creates so much variability that we cannot list prices here. Please send us your entire shelving system, including the controller(s), when possible.
From £1,250 +VAT — quoted per array. As mentioned above there are factors that affect the price of this service such as number of members, configuration of shelves and type of vendor format. Therefore this is scoped not priced. A 50% deposit is applicable for services provided and the balance is due on success.