In addition to Weibo, there is also WeChat
Please pay attention
WeChat public account
Shulou
2025-04-04 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > Servers >
Share
Shulou(Shulou.com)06/02 Report--
Hyperconverged infrastructure has been increasingly embraced by users and is considered the future IT infrastructure trend. Many users are already thinking about how to smoothly migrate production environments to hyperconverged architectures and how to build hyperconverged architectures around existing VMware vSphere ecosystems. While hyperconverged architectures are supposed to support multiple hypervisors, enterprises tend to build hyperconverged architectures around VMware vSphere for a number of reasons:
Compatibility of business systems with VMware vSphere has been fully validated.
We have trained a group of professional operation and maintenance personnel for VMware vSphere and have sufficient technical reserves. We are worried that replacing the virtualization platform will cost us a lot of learning.
There has been a lot of investment around the VMware vSphere ecosystem, such as VMware vSphere licensing, backup software, monitoring software, and even cloud platform integration.
So what are the considerations for sizing a hyperconverged infrastructure around VMware vSphere? Since the Hypervisor has been specified, the key is the selection of distributed storage software, so at least the following factors need to be met:
It has good compatibility with mainstream VMware vSphere versions, ensures the stability of the entire platform, and has continuous development capabilities.
Optimized for VMware vSphere runtimes to deliver superior I/O performance.
To provide online scale-out capability, the process needs to be simple enough.
Because it is a hyperconverged deployment, distributed storage software needs to ensure that resources are as low as possible under the premise of normal operation, so as to ensure that more resources are provided for virtual machines to use.
It provides advanced storage functionality support, such as storage synchronization and asynchronous replication, to meet the business continuity requirements of production environments.
SMTX Distributed Block Storage (ZBS) Hyperconverged infrastructure built on vSphere
Introduction to SMTX Distributed Block Storage
SMTX distributed block storage module (internal code "ZBS", hereinafter referred to as SmartX ZBS) is a high-performance, highly reliable distributed block storage system independently developed by SmartX. SmartX ZBS has been designed for hyperconverged architecture since its inception. Taking full advantage of the characteristics of virtualization platforms and SSD devices, in-depth optimization and integration have been carried out. Support hyper-converged platforms combined with KVM, VMware ESXi, XenServer and other virtualization/cloud platforms.
SmartX ZBS has the following functions and features:
high performance
SSD devices as cache, make full use of SSD low latency, high IOPS characteristics;
Support hyperconverged architecture, compute and storage run on the same physical server, storage is compute-aware, most I/O can be done locally, avoiding network latency.
high reliability
Ensure data reliability and business continuity during server downtime through multiple copies;
Multiple copies are written with strong consistency to ensure data consistency.
low consumption
Compatible with commercial x86 servers, no special hardware requirements;
Storage consumes only limited computing resources (6 vCPU 16GB memory), avoiding the purchase of additional storage servers and perfectly adapting to hyperconverged architectures.
horizontally scalable
Synchronous linear scaling of performance and capacity;
A single cluster can support up to 255 nodes.
architecture
SmartX is based on VMware vSphere hyperconverged deployment, mainly by deploying VMware ESXi on bare metal servers, while SCVM (SmartX Controller Virtual Machine) is deployed on VMware ESXi. SCVM, as the name suggests, acts as a storage controller virtual machine. It directly manages the hard disk resources (including SSD and HDD) of the server through passthrough technology (minimizing the performance loss through ESXi virtualization).
Multiple SCVMs are interconnected by a 10 Gigabit network, consolidate storage resources through the built-in SmartX ZBS distributed storage software to build a unified storage pool, and then mount to VMware ESXi using NFS protocol to provide storage services.
SMTX Distributed Block Storage (ZBS) combines the benefits of vSphere with user value
1. Good hardware compatibility makes user choices more flexible and stable
ZBS runs completely on the application layer, directly calls CentOS native drivers, and is compatible with most hardware on the market. It can effectively alleviate the problem of selection and stability risks caused by vSAN requiring independent maintenance of hardware drivers and inconsistent with vSphere HCL.
2. Loosely coupled with Hypervisor versions reduces user upgrade costs
ZBS adopts SCVM deployment mode under VMware hyperconverged architecture, which makes ZBS and vSphere version adaptation more flexible, reducing the additional cost of vSAN upgrade in native solution.
3. Data localization for better performance and business experience
In addition to SSD cache acceleration, ZBS also optimizes replica placement policies specifically for VMware vSphere virtualization environments. ZBS protects VM storage data through redundant replica technology, while data localization enables one complete replica to be accessed on the node where the VM runs, and the other replica to be stored on a remote node to ensure data redundancy. The benefit of doing so is that VM access to data at this node shortens the I/O path, effectively reduces latency and reduces pressure on network bandwidth, resulting in substantial improvement in I/O performance.
At the same time, ZBS perfectly supports VMware vMotion. Even if a VM is migrated online, the data localization function is still valid. The system will automatically detect that a VM has remote I/O access due to migration, and if it is not a temporary migration, the data migration will automatically occur and the data localization will be formed again on the new node.
4. Complete high availability mechanism ensures stable operation of user services
ZBS supports VMware HA functionality for node failure scenarios, and I/O route redirection functionality for other hardware failures, providing more granular high availability than HA.
(1) Storage network failure response
SmartX Halo supports automatic redirection of storage network links to business networks in case of storage network failure, temporarily serving as a link for storage data transmission without triggering VMware HA, and virtual machine services on hosts will not be interrupted. In the same case, VMware vSAN storage network failure directly triggers VMware HA to forcibly shut down the virtual machine on this node and restart the virtual machine after migrating to other nodes. In this way, business interruption and migration will occur.
(2) SCVM fault response
ZBS and VMware ESXi are loosely coupled, which makes HA mechanism more flexible. Even if SCVM fails to provide storage service, and ESXi itself does not fail, SmartX will trigger I/O redirection function, so that the virtual machine of this node can continue to complete by quickly switching the I/O request of virtual machine to SCVM on remote node. In this way, virtual machine service can continue to provide storage service without interruption.
5. Fast cloning mechanism effectively improves business efficiency
ZBS supports COW copy-on-write snapshot and clone technology. When snapshot and clone operations are executed, real data replication will not occur. Data replication will only occur when data is modified. Therefore, clone execution speed is very fast. Clone can be completed in a few seconds, and clone time is not affected by VM storage capacity. At the same time, ZBS supports VMware VAAI, which can offload VMware snapshot and clone operations to ZBS bottom layer execution, so that VM cloning through vCenter can also apply ZBS fast cloning technology.
6. Disaster recovery provides higher service continuity and data reliability
ZBS supports the deployment of synchronous replication clusters across computer rooms based on VMware vSphere, which can achieve "active and passive" in conjunction with application environments. It can also support asynchronous replication at the bottom of the storage layer. When the main computer room fails, it can be switched to the standby computer room to recover the business.
Visit SmartX's official website to learn more about deploying hyperconverged architecture solutions based on VMware virtualization platforms: www.smartx.com/solution
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.
Views: 0
*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.
Continue with the installation of the previous hadoop.First, install zookooper1. Decompress zookoope
"Every 5-10 years, there's a rare product, a really special, very unusual product that's the most un
© 2024 shulou.com SLNews company. All rights reserved.