PRODUCTS
*Registration of domains
*Hosting
*Web Hosting
*Dedicated server
*Housing
*Electronic Commerce
*DNS Administration
*voIP Telephony
*Control Panel

IDE vs. SCSI and RAID
Even though we are aware that the major limitation always is the velocity of the own communication on the internet, and that no matter how fast these are, they are never even close to the velocities of the transfer of data processing within the server and the local network, one of the objectives is saving the waiting time in all the processes.
The mass storage devices uses nowadays are set up by hard discs, which are mechanical and magnetic storage and recovery elements for the information.
In the last years the performance of the hard discs with interface known as IDE (this should really not be called for an interface it self) has developed very quickly, but even so, the most expensive interfaces with bus SCSI (Small Computer Standard Interface) exceed the performance of any IDE device by far.
In systems with high transfers from and to discs, even though this is carried out in the intermediate memory on discs (also called swap or exchange memory) it is preferable to use servers who incorporate hard discs with bus SCSI.
SCSI not only allow a high performance, nowadays with transfer rates given by the manufacturers around 160Mbps, but also allow other important advantages, from the own system for positioning the magnetic head for reading/writing on a disc, passing through the own communication language of the device, until the possibility of having units which can be connected and disconnected from the system without having to stop the operation of this (Hot Swap).
As the last characteristics which are applied to controllers of SCSI interfaces, we find the RAID systems (Redundant Array of Independent Disks).
There are five types or classes of RAID, but without wanting to get in to deep, lets say that some of them are focussed on the duplication of the channel for transferring data, its to say, provide the data at the same time from two devices with the resulting raise in velocity, and in other cases the RAID systems are configured to maintain the duplicity of the data in two volumes, physically independent as a security measure.

PERFORMANCE OF INTEREC SERVERS

All the systems from the level of the equipment to the communication lines, passing through the own switches and network distributors and all the racks, are designed and calculated with exceptional performance factors, as well as tolerance ranges and over sizing in many cases over 400%.

racks The performance of the Individual Server level, result from the services and the load of the server, according to its processing capacity (CPU) and memory capacity.
In this aspect INTEREC provide even multi processor platforms with kernel specifically compiled for the same, systems of great memory RAM capacity and arrays for hard discs Ultra Ultra Wide SCSI with transfers of 160Mbps in modes RAID-2 and RAID-4.

At this level we can value the Performance Management Policy for systems, in hardware and software.
In the first of these, the level of hardware, all the used elements are not only of high reliability with quality controls of maximum requirement, but also oriented at the maximum performance.

This way the used elements, pass a quality control phase, using only approved elements in the Servers forINTEREC for this purpose, among which we have motherboards made in USA by SuperMicro, IBM y Fujitsu hard discs, of very high performance and reliability, with the highest MTBF rates ( Middle Time Between Failure ).

Not to forget the racks that support these servers, the majority racks of 19" with a height of 4U, which facilitate the ventilation from a suitable ventilation system, the processors in these hardware are set at temperatures below 40 degrees, even in systems which have many SCSI elements for high heat radiation inside.

All this together with the a selection of the software of lowest level, its to say, closer to the language of the processors and compiled according to the experience and the measures about it self, makes the server units of very high power..

The objective of achieving the maximum performance is completed by a sufficiently over dimensioned network, using for this switches 100Mbps i the network communications, flows when it is possible towards circuits of 2M in Frame Relay instead of channels to ATM, and other techniques which result in a benefit for the total performance of the systems.