Subject: Better Data Backup Technique Required
We are currently using SQL server cluster for the storage of our financial transactions. And mirrored drives for main servers. I feel that mirrored drives storage technique is not enough for completely secured protected data backup of our media servers, developers’ workstations and desktops.
Using mirrored drives just save our data from disk failure and nothing else like fire, theft, power surges, sabotage, virus corruption, and accidentally changed or deleted files; as all the disks are right next to the system. Even in the case of a disk failure the server needs to be shut down for the time till the IDE1 is installed into IDE0 slot. Data could only be recovered if it is properly mirrored onto the second disk. However, the suggestion of SAN supports data mirroring and helps the process. SQL server clustering does not fully support data mirroring. As when the server switches from one to another the SQL gives an error message saying that the session was terminated and then the entire transaction has to be re submitted or even the transaction is lost. This is a loss of time.
I suggest storage area network (SAN) and backup for our media servers, developers’ workstations and desktops. In this we will have remote computer storage devices such as tape libraries and these will be attached in such a way that it will look locally connected to the servers. It could be used for remote locations for securing data from theft, fire and others.
SAN would give us the following benefits:
– Servers itself boot from SAN.
– SAN also reduces downtime, increases data security and fast access.
– SAN enables storage replication. By it a secondary storage at a distant could also be easily accessed by SAN’s extension over IP protocols.
– Consolidate servers and storage and then use SQL server clustering over it for better performance by mapping the storage applications on different clusters.
– Our video tutorials at some point require editing and this requires huge amount of data to be transferred. Therefore video groups require a prioritized bandwidth throughout the process and MetaSAN enables this functionality.
– Data could easily be migrated from one data storage device to another and enables data sharing among different servers.
– SAN as consolidation of resources results in cost-effective management, increased availability of data resources and better security, and even significant cost cuts.
– Interconnection between different forms of storage would be possible and easy.
We are not using any storage device or technique for images, and videos storage apart from mirrored disks. SQL server is only recording our financial transactions. We need a proper backup system for media servers and developers’ workstations too. As power failure might not be able to bring the lost data on developers’ workstation before it was copied on to the second disk. We give online video tutorials.
If no backups for media server then following problems may arise:
– If the server fails the streams from the failed server needs to be reassigned to a functional server in less time so that the service is nit disrupted much.
– If there is high customer interactivity, then the server might be overloaded and suffer from transient periods.
– If data backups are on a media server (RAM disk enabled) instead of on separate then the restore data files may be over-written after a system reboot and file system corruption. This may cause the server to become a failure and no more useful.
If our media server fails due to any accident then we will be able to recover our data through backup and our on-line tutorial service would also not be much disrupted by the failure. This is important as we do not want to annoy our customers and want to retain them.
The desktops and developers’ workstations need to have a backup too. Developers’ save essential and vital details on their systems and even a single file loss may be result in a huge idea loss. Entire work and hard work would go in vain. If we have a backup for these we would be able to save all the work and for a longer period of time for future referencing as well. On the desk tops, we are restrained by the storage capacity and mirrored drives disadvantages we have already looked at.
If we are able to deploy the above mentioned techniques, we will have a database created by our backups and we will be in a better position to analyze our customers, our developers’ and do future predictions, come up with improved solutions for our daily problems [if any occur].