What is QSM 2.0?
QSAN storage systems are based on the foundations of quality hardware and advanced storage operating system. QSM (QSAN Storage Management) operating system is composed of embedded Linux and in-house fine-tuning 128 bit ZFS file system. QSM 2.0 provides the best file consistency and data protection among others. Many enterprise storage functions are supported to help you consolidate file servers, increase storage efficiency and productivity, and ensure business continuity with high availability.
Powerful ZFS file system
ZFS (the last word for file system) is the ultimate, state-of-the-art, modern file system. It’s a transactional, copy-on-write, crash-free file system with 128-bit huge addressing space. ZFS can detect silent data corruption and has self-healing capability and guarantees end-to-end data integrity. There are abundant RAID levels to choose from to meet application requirements such as redundancy and performance. For comparison with Btrfs, please check here.
Storage pool architecture
ZFS supports storage pool concept to provide better scalability and higher efficiency to manage storage space. The conceptual architecture is shown below. A storage pool is comprised of many RAID sets. A RAID set is a group of hard drives with RAID protection. Usable space is allocated from the storage pool for host applications.
There is zero RAID initialization process and no formatting process. Once storage pool is created, it is ready to be accessed and support all kinds of data services. Check out our Youtube video to see the differences at https://www.youtube.com/watch?v=OkHmVGoM0Ks .
RAID member disks across enclosure boundary
When creating RAID sets, the member disks are allowed to come from different enclosures. This gives IT managers the maximum flexibility to allocate hard disk resources.
Hybrid storage and SSD caching
By default, ZFS file system supports SSD caching to take advantage of the strengths of SSD drives. Both SSD read cache and SSD write cache are supported. L2ARC acts as read cache between memory and hard disks. Frequently accessed data (hot data) will be moved into L2ARC to speed up random read performance and lower I/O latency. Deduplication can also benefit from L2ARC by moving the deduplication table into L2ARC for faster access. ZIL (ZFS Intent Log) acts as write cache. ZIL is a logging mechanism to conform with POSIX file system requirements for synchronous writes. For more information, please consult Qsan application note QAN201203-How_to_Deploy_SSD_Cache_in_QSAN_Unified_Storage.pdf .
Enterprise storage efficiency
QSM2.0 supports thin provisioning, deduplication, and compression functions without extra license fees. These three functions are designed to cram more juices out of the same storage space to increase more consolidation and utilization rate. Overall, they can help achieve 3x times more efficiency.
Storage pool locking
To avoid hard drives been stolen and add extra measurement of data protection, Storage pool can be encrypted at creation time. The encryption key can be exported to USB drive for future authentication needs.
Centralized file storage for cross-platform access
QSM2.0 can act as a consolidation server that integrates SMB3.0(CIFS), NFSv3/v4, AFP3.4, FTP, and webDAV network protocols. Heterogeneous hosts such as Windows, Linux, Unix, and Mac can put their files on QSAN NAS storage as a centralized depot for convenient file exchange. You can get rid of individual Windows file server, Unix file server, FTP server, or Mac file server and consolidate into only one unified NAS storage. It’s easy, convenient, saving hardware costs, saving electricity, and reducing storage management complexity.
Q-Turbo offload engine
Q-Turbo offload engine includes both SOE (Samba Offload Engine) and ZVIOS (ZFS Volume IO Scheduler) modules, which are two QSAN innovated in-house IO acceleration technologies. By adopting coalescence algorithm, SOE can greatly improve SMB (Server Message Block) protocol performance. ZVIOS can accelerate block access IO such as iSCSI and Fibre Channel protocol performance.
In addition to standard Ethernet network support, QSM2.0 can deliver iSCSI target and Fibre Channel target services in one chassis enclosure. It can help you further consolidate storage deployment with higher efficiency and smaller energy footprint.
Complete network bonding modes
QSM 2.0 supports all seven bonding modes, which include Round-Robin, Active Backup, Trunking, Broadcast, LACP, Transmit Load Balancing, and Adaptive Load Balancing. Hash-type setting is also provided. Please examine carefully with your network environment and choose the correct setting.
Advanced network capabilities
To cope with complex enterprise network environment, QSM 2.0 is equipped with static routing and diagnostic tools such as loopback, traceroute/ping, and ARP (Address Resolution Protocol) functions.
Support Windows Active Directory
QSM 2.0 provides the best integration with Windows Active Directory service and Windows ACL function. You can effortlessly deploy TrioNAS system into Windows working environment. Both local user accounts and domain user accounts can co-exist to allow the maximum flexibility and easy management.
Support LDAP Directory service
QSM 2.0 supports LDAP protocol as well for Linux and Unix environments using LDAP-based directory service such as OpenLDAP to simplify account management.
UnifiedAUTH for unified authentication
Imagine the working environment with Windows server, Unix server, FTP server, and iSCSI target storage. You probably will need AD account for CIFS service, Unix account for NFS service, FTP account for FTP service, and iSCSI CHAP account for iSCSI target storage. Try to memorize all these accounts and passwords every day. UnifiedAUTH releases you from this shackle by using the same account authentication for all data services including iSCSI target! You only need to remember ONE password!!
Dual active controller design
Enterprise applications require stricter demands for service availability, which is a measure of system up-time and is expressed as a percentage of UP time / (UP time + DOWN time). For enterprise storage array, the requirement of availability is usually 99.999%, which translates to about five minutes of downtime in a year. Both storage controllers can deliver I/O services at the same time. Firmware updates are transparent with zero downtime and no business disruption. Active-active implementation can scale performance up to 1.3 times and double controller memory capacity to increase the cache-hit ratio.
In case error or malfunction happens to either storage controller, the peer controller will resume all data services and tasks from the failed controller and becomes the master controller. Inter-process communication, cache mirroring, and heartbeat between the two controllers are carried out through cable-less, high speed bus (64Gb/s) on backplane board. Failover time is around 15 seconds and less than 60 seconds on average. It greatly minimize the window of vulnerability of service-sensitive business applications. Check out our Youtube video demo at https://www.youtube.com/watch?v=XmeuZ1SdJmU .
Fail safe networking
In addition to controller fault tolerance capability, the same protection extends to data access paths, which comprise of mainly network ports, switch ports, and network cables. Fail safe networking can tolerate the failure of any component in this data access path thanks to cluster IP technology, which is composed of mirrored network ports from both RAID controllers. So each cluster IP contains dual network paths, one from each RAID controller. Should any network component along the path fails, the network packets will be quickly switched to the redundant path. The transition is TRANSPARENT. SMB and NFS file access can continue as if nothing happens at all.
Online firmware updates
To provide the highest level of system availability, QSM 2.0 supports zero downtime for firmware updates. After updating the firmware, dual controllers take terms to reboot. When one controller reboots, all the data services and tasks will automatically failover to the peer controller. Your business applications can keep running without disruption.
File snapshot is basically taking a snapshot against a “volume”, which is the predecessor of a shared folder. Everything inside the volume will be protected by the snapshot based on block-level, copy-on-write technology. In case you accidentally delete a file or make a wrong modification to a file or files under virus attack, you can easily retrieve the file with a few mouse clicks.
LUN snapshot is basically taking a snapshot against a LUN, which is used by iSCSI connection or Fibre Channel connection. Using the same copy-on-write technology as file snapshot, differential data are recorded between two given points in time. You can rollback the snapshot to a certain point in time in the past or clone the snapshot to become an independent LUN.
For file snapshot, QSM 2.0 implements auto-mount function that automatically mounts each snapshot to its target volume. In Explorer of Windows OS, each snapshot appears as a hidden folder. You can effortlessly retrieve the file at a given point in time in the past by clicking inside the snapshot folder. No hassle at all!
Disaster recovery – local clone & remote replication
By leveraging its own snapshot function, QSM 2.0 supports block-level, snapshot-based asynchronous remote replication. Only differential data (snapshot) will be replicated to the remote site, so it is very efficient. Both volume and LUN can be replicated to provide the full range of protection. Scheduled tasks are supported for easy management. QSM 2.0 can save you thousands of dollars for deploying host-based replication and free up host server resources.