The new life of the old storage - Violin magic iron to speed up arrays

If you have such storage systems as EMC Clariion, VNX, VMAX, Symmetrix DMX3, DMX4, AMS 2000, HUS and others like that, and you lack their performance, I have good news.

You may not need to buy a new fast storage system. If you just need to speed up the reading tasks, there is a solution that is much cheaper than upgrading the array and easier to implement than disks in Symmetix. It is called Violin Maestro.

This is a hardware read cache that connects between the host and storage. The piece of iron is already tested and already in Russia. It can be taken and set up without any downtime and stops.

What and how

Actually, the principle is very simple. Violin Maestro is put in the gap between the store and the host and caches data. Data is sent from the caching device, and it works at speeds close to DRAM.

You can speed up both new arrays and old ones - the cache does not matter which array to work with. The only thing is that the sequential read task class does not get into the cache. Maestro is smart and does not touch a consistent load. The whole main butt for working with databases, virtual machines is accelerated just great.

The implementation is very simple, and is done without downtimes. First, Maestro is put in bypass mode and simply scans traffic that runs past the device. He studies, analyzes the data, looks at the use of each moon and considers the optimal cache mode for them. Then, after training (in practice, a day later), he recommends the optimal mode for each volume. What is nice - among other things, it calculates the shift (when the blocks of the necessary data begin from the middle of the storage system). All this can be configured through the GUI and through the console with your hands.

Further, the device pulls hot data into the cache and responds to host read requests on it. If there is no data in the cache, the device simply passes this request to the storage. Additional delay - just no more than 300 μs. The host does not see Maestro - for him it is, roughly speaking, a "transparent client", something like an extraordinary FC switch.

Despite the fact that Maestro speeds up reading only, during test implementations we noticed another interesting fact. When part of the read load is removed from the storage system, resources are released that make writing faster. That is, if you count in the performance metrics of the butt working with "hot" data - the whole system is accelerated.


This is what the Violin Maestro array looks like

Implementation

Here is an example of implementation in a pharmaceutical company, where about 4 thousand users worked in a system with VMware / NetApp (it was on the left and on the right):



The result is this:

• I / O latency (average latency) decreased by 8 times.
• The number of IOPS has grown 12 times.
• SAN utilization increased 4 times.
• Pharmacists saved $ 1.2 million compared to installing a new storage system, and later became so involved that they saved another $ 775 thousand when they realized that 3 times more users could be placed on the same hardware.

It is worth noting that Maestro is not a panacea for speeding up storage. Here are the main situations:

• If your reading tasks prevail, Maestro will help with hot data cache. In this case, the decision will cost 10 times cheaper than buying a faster storage system for reading.
• If there are a lot of write tasks, Maestro will free some of the resources by processing read requests before storage, but it will not fundamentally change the situation. As a rule, you will need to upgrade the array.

Basically, of course, Maestro is being introduced where the budget does not allow you to buy a new Hi-End storage system, but you need a quick read. Therefore, the usual practice is a good disk storage + Maestro, or a good old Hi-End + Maestro. Examples of integration include insurance, banks, universities, medicine, the oil and gas sector, as well as retail, telecoms, and so on. All where something slows down.

TTX for model 2510

Reliability

In data centers, a piece of iron can be considered reliable only when it has not presented surprises for several years in a row. For the storytellers, the fact that the device was “good old” is recognized at the time when it is time to remove it from the rack and write it off, because the lifespan of an ordinary storage system is about 5 years.

A clear concern about Maestro is that this is another link in the chain between the host and the storage. I will say that I personally trust the piece of iron. Firstly, of course, there is no single point of failure (this is the norm for the data center hardware), and secondly, in case of problems there is a bypass that allows the host to simply work with the storage directly. Thirdly, updating and another service is done without downtime: for example, a new firmware - like a Martian lunar rover. First, the software receives one controller, reboots, if it is successfully operational, the second starts to update. Fourthly, volumes can always be returned to hosts by giving directly from the storage system - the system does not make changes to the data on the storage system, which means they are safe.

About the bypass, by the way, I must say separately. It can be done selectively. We had an example when 5 test bases and 5 bases of the “combat” butt on production were accelerated. At the end of the month, a very resource-intensive task arose in production, admins simply switched test bases to bypass mode, which allowed us to give more cache to the main system.

Interesting?

It was damn interesting for me to hold Maestro in my hands and see what a piece of iron can and what cannot. I must say, the final impression was even slightly better than I initially thought when I heard the “smart cache”.

I do not propose to take my word for it. I suggest to take and try. This can be done at a conference in September, where new Violin solutions will be presented. There you can discuss test drives for your company.

Conference

If you work with IT infrastructure, then I invite you to a conference on new solutions for data centers. There will be guys from Violin (including engineers and tops), our engineers and data center specialists, as well as representatives of customers from Ingosstrakh and Raiffeisenbank who have already tried the hardware from Violin.

The program
includes : • A detailed story about Maestro (detailed, highly detailed and with more details than the above).
• Windows Flash Array storage that supports RDMA (ideal for Exchange, SQL Server, Hyper-V).
• A new virtualizer, which is very necessary if you have several storage systems. More about him is already in place. For CIOs, the opportunity to cut infrastructure costs next year.

Come, we’ll show you everything. September 25th, here is the linkon the program of the event and registration. Well, traditionally, you can already ask me about Maestro and implementation by e-mail vbolotnov@croc.ru. Or in the comments. Just keep in mind that we have only one test piece of hardware, and for the next two weeks at least it has already been scheduled for tests.