Shulou(Shulou.com)06/01 Report--

Many novices are not very clear about how to interpret the EMC SYMMETRIX VMAX IWeibo process. In order to help you solve this problem, the following editor will explain it in detail. People with this need can come and learn. I hope you can get something.

I won't write too much today. Let's talk briefly about VMAX's iUnip O process.

Why are we talking about the IWeiO process? Because the architecture of high-end storage is more complex, especially in the case of distributed CACHE like VMAX, many people do not understand what it is like to read and write VMAX with so many nodes.

In fact, it is not as complicated as everyone thinks. Although the VMAX Cache is physically distributed, it uses a logical unified addressing method, so it is logically a global Cache, but the Cache accessing this engine is local and faster, while the Cache accessing other engines needs to go through the Rapio IO switch with greater latency. But from the point of view of the large process, there is no difference between VMAX's Icano process and DMX, which is also the advantage of EMC. The physical architecture has changed, but the Icano process remains the same, and a lot of code does not need to be rewritten, saving a lot of research and development costs.

This is the process of reading the hit, which is relatively simple. The host sends a read request to VMAX,VMAX and finds that there is this number in the Cache, then returns it directly to the host and modifies the metadata. Here the metadata should be used to manage the CACHE algorithm, including the number of cache hits, the time of the last access, and so on. The upper algorithm needs this metadata to determine the action of Cache updates.

This is the process of reading without a hit. Since the read information is not found in the cache, the back-end interface is needed to read the corresponding data from the corresponding hard disk, read it into the global cache, and then return the data to the host to modify the metadata.

Then let's take a look at the writing process. If VMAX's Cache is plentiful now, VMAX is following a fast-writing process. The host initiates a write request to the VMAX, and the VMAX is written directly on the Cache, and then tells the host after updating the metadata that it is finished. Finally, the backend interface sets the disk asynchronously according to the set parameters, that is, it is not necessary to set the disk every time, but generally reaches a certain threshold. It should be noted that because the write operation is to prevent the loss of power-off data, it is necessary to update the metadata to tell the host that it is finished. This is different from the order of read operations. Because cache is mirrored, there are two copies of data and metadata to avoid a single point of failure.

When the Cache of VMAX is almost full and exceeds the set threshold, such as 80%, then the data in Cache must be unloaded before writing to Cache, and then the metadata must be updated, and the subsequent process is the same as when writing. This is the delayed fast writing process.

As you can see, for ordinary users, VMAX is a whole, regardless of the communication between engines, VMAX will deal with it internally. The actual process is much more complex than this, but the average user does not have to relate, because Cache is logically already a global Cache.

Is it helpful for you to read the above content? If you want to know more about the relevant knowledge or read more related articles, please follow the industry information channel, thank you for your support.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.