Before the deployment of Zabbix, we experienced some delays on our internet connection, we decided to deploy it and we registered all our network devices, such as switches. We turned on the firewalls in Zabbix. We linked Zabbix to Grafana to receive a more accurate graph.
From the use of Zabbix, we are now able to see an overview of our bandwidth consumption, a direct view of what links are being used on our backbone switches, and if there is any unusual peak of traffic on our network devices. Before using Zabbix and Grafana, we didn't have any view of the results of our network. We used to connect individually to every single device and collect the data to analyze but now, we have the opportunity to see graphs and charts of all of the network metrics.
I have found that the reporting feature in Zabbix is most valuable. Additionally, the solution has given us bandwidth options, we are able to see where problems are. For example, we noticed a problem that occurred because of a bad interface going in the wireless VLAN.
The user web interface is a little bit too basic, we need to link Zabbix to Grafana to have more options, such as graphs and charts. The interface needs to be improved. Additionally, there could be better integration with Grafana API.
I have been using Zabbix for a couple of months.
Zabbix is very stable. The resource consumptions are okay.
I cannot give too much information about the scalability features because our test environment is not very large. We have a small environment of only some firewalls and switches. We have some servers also, but we did not register them on Zabbix.
We have two users that are using this solution, the CTO and the principal network engineer.
We do not have plans to increase the solution right now but in two years we will.
The initial setup is straightforward.
I only needed to download the application and then deploy it in the VMware Workstation, that was it. Once the virtual appliance is connected to the network, it will automatically receive access via DHCP. Lastly, we use the interface to configure specific ports, the initial deployment is quite simple.
The time it takes for the deployment takes only a few minutes.
We only require one network engineer for the maintenance of the solution.
We have received a return on investment because the solution is free and any gains we make are profit.
Zabbix and Grafana are both open source products, we only needed to go to their website and download the application and we began to use them. The solutions are free.
Zabbix is a good and very cost-effective product. It does its job very well. We downloaded it for performance monitoring and it's very effective.
I rate Zabbix an eight out of ten.
Hi everyone,
I want to share a story about cooking toasts and tell you how we were expanding the functionality of Zabbix by a coil of wire and an integrated circuit Open Source solutions. First things first.
Imagine, you work in an organization (and it can be yours) where its business process is supported by one or more IT systems. There is a monitoring system. Then a little blurred vision, and it is not clear whether /the industrial system is free or Open Source. However, there aresituations when all the sensors aregreen, but the business process gives unexplained setbacks, showing a decrease of key indicators. Matter-of-factly, the situation is out of control and your intervention is required. It’s quite a common case. A systematical approach for solving this problem will provide a drafting services model.
Tom Wujec who provides services for the visualization of the processes occurring in the companies (not just IT), carried out a curious study. He asked different people to draw the process of toasting. See below for some results of this work.
What do we see on this pictures? Right you are. You can see objects and links presented in any system. The more you see them , the more the approach will be systematical. The correct degree of granularity will more accurately monitor the "health" of business system. To construct a circuit, you can try to use the Visio, but much more interesting is to take a marker to draw connections, adhesive sheets for objects and to portray your system on the whiteboard. The more is the quantity of yellow leaves, the more are the links, the higher is the chance to determine the maximum number of monitoring points for right resolving of the problem.
And now it's high time to tell you about our experience in the field of expansion of standard functional Zabbix and of systematical approach described above. There are exactly two of them.
The first is Mapping systems and the installation of thermal services’ card. In view of our considerable experience in the monitoring of the banking business processes, we give an example of exactly this area. Let’s consider three of the most common banking systems. If there is a more typical system in your bank it will not be considered here.
Home banking (online banking):
Enterprise serial bus (ESB):
Decision-making system (Credit convey):
In our application, it will look like this (yes, the structure is somewhat broken, but visibility still persists):
You can go to a lower level in case of need. And, last but not least, when you move the object pops up pop-up window with the description of the event and with a link to the graph in Zabbix
Through this approach, and given the level of details, we get friendly dashboards in the output, and at the same time a simple problem of localization tool. A few words about the features of our system:
- Visualization of the dependencies among enterprise systems;
- Set up of the degree of mutual components’influence (due weight);
- Integration with Zabbix (objects on the heat map are associated with flip-flops);
- Pop-up windows with text events when you hover over the event object;
- Visual Interface Communication Configuration objects;
- Visual communication setup interface of objects with triggers Zabbix.
As an example, some already developed interfaces.
Adding objects to a heat map:
Adding integrations from Zabbix:
Connecting Zabbix triggers to objects on the heat map:
The system is based on Google Charts and Bootstrap. While this is an alpha version, we plan to develop it further, adding useful pieces of industrial systems, which have been working successfully for many years. I will try to keep you in touch and to publish posts on the basis of practices developed.
The second experience is the integration with Zabbix and heat maps of synthetic transactions functionality. In fact, it’s a continuation of the heat map, and the look from the other side. Definitely controlling the system only from the application and infrastructure, you will not have the necessary information completeness. Synthetic transactions will allow having a look as a user and isolate the problem even before the first user requests to the Help Desk.
Synthetic transactions are based on the framework phantom.js (but nothing prevents you to go to casper.js, to pure selenium or anything else on your liking). In our test lab the execution of scenario is configured through cron and then the data are transferred to Zabbix by Zabbix trapper. As an example, the test scripts are logged into the MTS (Russia's mobile operator) personal account and receive money from the demand balance and the traffic in the Internet package (I can send it to you if you interested in this script). In the banking community, the most likely use of this tool can be, for example, online banking. Nobody will prevent you to log into the system and flip 1 dollar from one account to another.
Collected items from our example are as follows:
Each has its own schedule.
I do not want to say that the use of Open Source solutions monitoring is a tablet from all troubles. I will reveal an open secret: as in physics, there is the law of conservation of money and effort. The more money you invest into a finished product, the less effort for improvement, and vice versa. Always should be guided by common sense, the available budget and the human factor: ready to be your team to throw in the recess of business monitoring on the first call?
The Zabbix is top because is fundamental easy to manager.