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

This article will explain in detail about EMC Symmetrix SRDF analysis and discovery, the content of the article is of high quality, so Xiaobian shares it with you as a reference, I hope you have a certain understanding of relevant knowledge after reading this article.

Today, let's take a closer look at SRDF and see what features it has in addition to its rich networking topology.

Let's first look at SRDF's support for N:1 replication. I did not find the relevant parameters, but I saw that Enginuity limits the number of sessions to 16, so in theory SRDF supports up to 16:1 centralized replication.

Second, let's review the concept of Symmetrix logical volumes. As mentioned earlier, Symmetrix logical volumes have up to four mirror locations, which is the biggest difference between EMC's mechanism and other vendors.

We see that local RAID takes up one mirror bit, so there are only three mirror bits left, and SRDF also requires one mirror bit per replication connection, which limits SRDF to support up to one logical volume replicated to two devices simultaneously. Because hot spares are required for RAID groups in general, hot spares also require (if Direct Sparing mode is used instead of Permanent mode) a mirror bit, and snapshot volumes BCV also require mirror bits, making Symmetrix mirror bits the most valuable resource.

This mechanism imposes more limitations on SRDF scalability, but it also brings a huge benefit. In other words, the mirror bit of SRDF represents a piece of data, which can be used to protect local data from corruption and share the local load without switching the host. For example, if there is a local hard disk failure, SRDF can read data from the remote and mitigate the impact of RAID rebuild on the system. If the promotion is opened, if a local double disk failure occurs, the data is gone, and the production host can also read and write data from the remote through SRDF without switching. We'll sync it back after we fix the hard drive locally. General manufacturers need host switch, the process is more complicated, because the site switch is easy not to do, the risk is very high, the last ICBC upgrade failure, a lot of netizens questioned why not switch, downtime for so long.

EMC's official documentation for RPO of SRDF/A only states less than 1 minute, or seconds.

But let's analyze it. SRDF/A adopts a similar way to Fujitsu and Huawei, a bit like ORACLE redo log, grouping I/O in order, and then replicating cyclically to ensure the order of replication.

If a failure occurs, the remote end loses approximately 1-2 change sets of data.

The minimum cycle time for this change group is 1S, which means that ideally only 1- 2 S of data may be lost, so an RPO of 3- 5 S should be possible from a mechanical point of view.

Of course, if the Cache space is not enough, these changes will be written to the hard disk to ensure that the link is not interrupted.

Synchronous and asynchronous SRDF can switch automatically on a case-by-case basis, resulting in less data loss.

Remote and local volume groups support various combinations of Thin and Thick.

Another important concept is SRDF/CG. This SRDF consistency group is mainly used to ensure the logical order of data. For example, multiple data in ORACLE database are on different arrays, while ORACLE I/O adopts the method of writing logs to prevent sudden power loss. Therefore, it is submitted in two phases, that is, writing logs first, then writing data, and finally writing logs. If journal and data volumes are on different arrays, SRDF/CG is required to maintain their logical relationship. There are some differences between this and snapshot consistency groups. Snapshot consistency groups can only guarantee that the data is the same time, and cannot guarantee that the data of journal volumes and data volumes match. These two functions are usually used together.

Online data migration is now supported in the latest Enginuity 5876 release.

Of course, a big improvement is that the same data volume can support both SRDF replication and Recoverpoint. In this way, the CDP solution of Symmetrix is more perfect. Other high-end solutions have not seen CDP solutions for the time being.

About EMC Symmetrix SRDF analysis and discovery is how to share here, I hope the above content can be of some help to you, you can learn more knowledge. If you think the article is good, you can share it so that more people can see it.

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.