English | Russian
Svrmon - fasted real-time server monitor

Appearance WEB interface

For monitoring and diagnosis of servers running on *nix platform there are many tools and resources, covering almost the entire range of possible tasks. Therefore svrmon is not unique in its kind, and only intended to simplify the task of setting a network monitoring system or cluster of a large number of servers.

Current service opportunities are quite simple:

So in the following illustration shows the appearance of the page display svrmon when entering it through the WEB interface - for 4 nodes with the names q-server, mysql, hiasm.com and localhost displays information about the CPU, available memory and loading drive for the root file system.

Set the monitor on a server running Linux

To install and run the monitor on the server (or workstation) running Debin base distributions simply connect to the repository from site hiasm.com:

echo 'deb http://deb.hiasm.com/ lucid devel' >> /etc/apt/sources.list
refresh list of packages:
apt-get update
and install package svrmon:
apt-get install svrmon
To verify that the monitor is running, to open in the browser URL address:
http://localhost:1718
To install the monitor directly from the *.deb package to download at http://deb.hiasm.com/pool/devel/s/svrmon/ correct version for you system and run next command:
dpkg -i svrmon_x.x.x_xxx.deb
For all other systems need to download the binary archive of one of the architectures:
svrmon-amd64.tar.gz (550Kb)
svrmon-i686.tar.gz (551Kb)
and run next commands with root privileges:
tar -xf svrmon-xxx.tar.gz
./install.sh

Configuring master server for the collect of data from child nodes

A necessary condition for a master server is the presence of fixed svrmon on all child servers, with necessary to obtain data on their condition. Furthermore it is desirable to install one and tighter version of the package as format of the output data may vary.

When all the conditions listed above, you must open the future master server configuration file subsidiaries nodes monitor, for example, in the editor nano:

nano /etc/svrmon/nodes.cfg
By default, the file content is as follows:
# nodes list in format:  
127.0.0.1 localhost
ie it contains only one node, pointing to the local server. To add an external server must list each of them in a separate line showing the IP address and name (the name can be anything). Example:
127.0.0.1 localhost
10.10.1.1 myserver1
10.10.1.2 myserver2
10.10.1.3 myserver3
10.10.0.1 gate
10.10.0.2 nginx frontend

After saving the list of nodes need to restart the daemon for the entry into force of the settings:
/etc/init.d/svrmon restart
To verify the configuration of nodes need to go to the following address:
http://localhost:1718/nodes
Page setup

Svrmon monitor allows you to create any number of pages with the correct set of output parameters and their formatting. To configure the pages to open the configuration file pages.cfg

nano /etc/svrmon/pages.cfg
By default, the file content is as follows:
/ root.html
/version version.html versions.html
/node node.html nodes.html
/data data.html
/os os.html oss.html offoff
The format is simple: the first is the page address, then the template name of the folder /etc/svrmon/template/ for display the local server (pagename), then the name of the template for the page with the parameters of all nodes listed in nodes.cfg and the last option - the template for the body page for not avaliable in currently node.

For templates of the local server and the connected nodes have two different sets available for inserting the page parameters:

Template surveyed nodes
NameValueStatic
_{*
nodesnodes hash map-
@|node_namename of current node-
@|node_bodycontent of request page-
@|node_ipIP address of current node-
Template localhost
NameValueStatic
_{*
cpu_idlecpu idle in %-
cpu_loadcpu load in %-
ram_loadram load (with cache and buffers) in %-
ram_totaltotal ram in bytes-
ram_freefree ram in bytes-
ram_bufferbuffers ram in bytes-
ram_sharedshared ram in bytes-
root_loadload drive with root FS in %-
root_totalcapacity drive with root FS in bytes-
root_freefree space on drive with root FS in bytes-
root_availavailable space on drive with root FS in bytes-
sys_load1system load at last minute-
sys_load10system load at last 10 minutes-
sys_load15system load at last 15 minutes-
uptimeserver uptime in seconds-
procsnumber of running process in system-
os_verversion and name of the distribution*
hostthe name of local server*
kernelversion and type of kernel*
versionsvrmon version*
scsithe array of serial numbers volume drivers*

* the "static" means the calculation of the value only once at startup daemon. All non-static parameters are updated at intervals of once per second.

Examples of ready-made templates for the pages can be found in the /etc/svrmon/template/

Additional information

Monitor svrmon written in Go for tracking the status of servers in the domain hiasm.com and of course to verify this works Google as a basis for future service, which would already fairly simple procedure for application assembly (in the language Go of course) to make even more simple, transparent and accessible to everyone. Because of its major advantages can be noted quite a number of powerful tools, delivered directly to the compiler, as well as high-speed assembly process of the program (compile and link).

As described in the review of the program is installed on all servers in the domain, then any visitor has opportunity to see the effect of the monitor in combat conditions, by clicking on the following link: http://mon.hiasm.com/cpu_logs

For all questions on the program refer to forum or at the mail of official support support@hiasm.com

create by dilma