Alongside storage technologies and network components (switches, routers), servers form the basic foundation of an information system. Servers house the computing resources needed to run applications. Servers are connected to the information handled by applications in a variety of ways: local disks, direct SATA/SAS/SCSI connection to storage systems, connection via SAN or IP network to advanced storage systems, implementing high-availability functions on the scale of a computer room, a geographical site, or several remote geographical sites.
SUPERVISION: WHY AND HOW?
Whether you’re a user of IT applications or SaaS/Paas/Iaas services, server availability and performance are essential to your satisfaction. That’s why it’s essential to monitor all server hardware and software characteristics.
WHAT ELEMENTS NEED TO BE SUPERVISED ON A SERVER?
Key points of server monitoring, hardware side :
-
temperature
-
component status (fans, power supplies)
-
disks
-
RAID systems
-
memory
-
processors
-
peripherals
Servers are equipped with redundant, hot-swappable components: it’s easy to see why, with continuous supervision of server hardware and a high-performance incident handling process, permanent hardware availability is an easily achievable goal.
Key points of server monitoring, software side :
- RAM usage
- swap
- storage space
- processors
- network cards
For detailed supervision of server performance, you need to monitor the length of queues at processor and disk level. Disk supervision is also relevant, as it is decisive in ensuring data access performance: read/write operation throughput and latency must be implemented as part of server supervision.
HOW TO SET UP SUPERVISION ON A SERVER?
Given the wealth of possibilities offered by server manufacturers, setting up server supervision can be a long and tedious task. The faster and more efficiently this can be done, the more the user of the supervision application will be able to rely on a complete catalog of service models, organized in the form of equipment models.
REPORTING
Know the status of your backlog in real time, and adjust your staff workload if necessary.
See in real time the evolution over 3 days, the number of requests opened and closed.
Know average resolution time, SLA compliance, etc.