SysOrb Network Monitoring System Administrator's Guide: For version 4.6.0 | ||
---|---|---|
Prev | Chapter 2. System Requirements | Next |
The requirements for the SysOrb Server depend on the number of NetChecks, AgentChecks and snmpChecks monitored. If only a small number of these need to be monitored the SysOrb Server can be run on very modest hardware. The more checks monitored, the higher the requirements for the SysOrb Server.
Please note: You would usually want to dedicate a machine to running the SysOrb Server, in order to minimize the risk of failures caused by other software, on your main monitoring Server. The SysOrb Server can, however, easily run on a system which is handling other tasks as well (such as a mail-server or name-server). It is not recommended to install the SysOrb Server on a system that is already very busy, as the SysOrb Server will consume resources, and some measurements (such as network response time measurements) will be disturbed and inaccurate if the system is too busy handling other tasks.
Tip: In general, the database space requirements will grow with approximately one megabyte, per monitored check, and one megabyte per monitored node. So monitoring 20 devices on each of 10 different machines, will require approximately 210 megabytes of disk space.
In addition to this, reports in particular will also add to the disk space consumption. Reports are stored efficiently and a fully detailed report over the aforementioned 200 monitored checks will require much less than the 200 megabytes required for the full scale monitoring. Yet, when setting up a system in which many users have access to generate reports, report disk space consumption should be taken into account.
One should, of course, configure a disk space check with suitable warning and alert thresholds, for the filesystem on which the database reside, when configuring the agent on the SysOrb Server.
Windows 7 or Windows Server 2012 (x86_64):
CPU: Intel Core2 compatible or better
RAM: 1 GB free memory for smaller installations (less than a hundred nodes configured). 2-8 GB recommended for larger installations.
Disk: 100MB for installation and log files and approximately 1 MB per configured check. E.g. an installation with 100 nodes and 100 active checks for each node would require 10 GB of disk space for the database
Red Hat Enterprise Linux 6 (x86_64):
CPU: Intel Core2 compatible or better
RAM: 1 GB free memory for smaller installations (less than a hundred nodes configured). 2-8 GB recommended for larger installations.
Disk: 30MB for installation and log files and approximately 1 MB per configured check. E.g. an installation with 100 nodes and 100 active checks for each node would require 10 GB of disk space for the database
Solaris 11 (x86_64):
CPU: Intel Core2 compatible or better
RAM: 1 GB free memory for smaller installations (less than a hundred nodes configured). 2-8 GB recommended for larger installations.
Disk: 50MB for installation and log files and approximately 1 MB per configured check. E.g. an installation with 100 nodes and 100 active checks for each node would require 10 GB of disk space for the database
Debian GNU/Linux 7.0 (x86_64):
CPU: Intel Core2 compatible or better
RAM: 1 GB free memory for smaller installations (less than a hundred nodes configured). 2-8 GB recommended for larger installations.
Disk: 30MB for installation and log files and approximately 1 MB per configured check. E.g. an installation with 100 nodes and 100 active checks for each node would require 10 GB of disk space for the database
There are many other factors that affect overall system performance. Large installations may need not just bigger but also faster storage. In general, more memory will help the disk I/O system. Using servers with more than one CPU core is also beneficial to SysOrb, especially in large installations.