Technical information

65
Reliability
– not many companies quote their reliability data, but this is where Fujitsu is different. Fujitsu is
committed to providing the most reliable products possible. The latest statistics show the teamserver Ci only
experiences one incident every 9.0 years and this is before the advantage of redundancy!
In addition our factories implement a “licence to build”, a rigorous training programme that ALL
manufacturing employees have to pass in order to build Fujitsu products, followed by continuous spot
checks.
Fujitsu reliability
Hardware availability figures based upon 4 hour call to fix/replace and Fujitsu hardware warranty statistics
(Nov 98 – Jan 99)
Ci 99.995% hardware availability (avg) - 1 incident every 9.0 years
Gi 99.987% hardware availability (avg) - 1 incident every 3.4 years
Mi 99.978% hardware availability (avg) - 1 incident every 2.0 years
The reliability figures quoted are in fact the worse case scenario, as no allowance is made for using
redundant or hot-swap/spare components. With these features incorporated the reliability figures would be
even better.
Of course, as servers get larger they contain more system components (e.g. disks) so the component failure
rate per system will increase. But the larger the system, the greater the impact there may be to the user’s
business. So redundancy coupled with hot-swap and hot-spare components become increasingly important
as systems become larger.
Manageability
While RAID, clustering, and other sophisticated features can provide a safety net should the unthinkable
happen, it would be much better if you could know about it before it happens, and so take corrective action.
Manageability provides an unprecedented level of control over systems - so much so that all the many
elements can be administered remotely and faults can often be identified before they occur.
Fujitsu
server
manager
Today’s systems are not getting any less complicated, and neither is their workload. Which makes it all the
more essential to manage these machines effectively.
This is where Fujitsu’s
server
manager
can help. It comprises a powerful set of tools, based around a
Windows-style interface, and delivers intelligent and reliable monitoring, alerting, remote maintenance and
control of servers on the network.
Fujitsu provides all of this free.
Fujitsu takes the view that manageability is not a luxury, but an
essential and integral part of any server.
As a result, the MIS department can keep track of its hardware assets, such as disks, cabinets and I/O cards
and will have a record of the configurations used. They can monitor the demands placed on the system, and
whether it has the resources to cope. And they will be kept informed of fan status and temperature,
providing an opportunity to identify and correct any problems that arise.
All of this can be achieved remotely, via a LAN or WAN or even via a modem, using an optional extra
known as
server
manager Assist
.
The benefits can be considerable, with an opportunity to reduce maintenance costs while increasing the
productivity of users.
Data Redundancy
- Lose the data or get corrupt data and the whole business might be in ruins. Data
redundancy is the most important form of redundancy to have. Therefore best practice organisations
regularly back up their data.
To ensure the very latest data is protected one further step is needed. Redundancy of live data, whether on
disk or in memory. The two most important forms of data redundancy are:
•
Error Checking and Correction memory (ECC)
•
Disk data redundancy - RAID
Both will correct the failure on the fly without impacting the server application or users.