US side: control-m（ www.bmc.com )
Chinese side: taskctl（ www.taskctl.com )
ETL scheduling tool Sino US pK (taskctl vs control-m)
But in China, I don’t seem to have any hesitation in choosing taskctl among many softwares. Although the software is not famous, I am very impressed by its fresh interface, unique design and user experience. I think, in time, taskctl will surely have its status in the world. Well, praise or less to say, evaluation technology to be objective, we still stand in an objective position to have a Sino US PK!
Let’s talk about PK methodThese two software both claim enterprise level scheduling software. We first PK the software enterprise level features, and then PK the software function points. Finally, the most critical thing of PK is user experience!
Enterprise level feature experience PK
To tell you the truth, I can’t define the enterprise level characteristics of scheduling, but there should be at least the following aspects: network support capability, cross platform capability, stability, large-scale data support capability, centralized data management, unified application portal, etc. Let me just compare them from these aspects.
1. Network support capability, which is mainly determined by the core network architecture of the software. The two softwares control the network in a multi-level way through EM node, server node and agent node respectively;
2. Cross platform capability: taskctl only supports UNIX linux environment, while control-m supports various mainstream operating systems;
3. Stability. It’s boring, but it can’t be avoided. Stability can’t be solved by software testing, but it still needs long-term test in actual environment. In this respect, taskctl cannot be compared with control-m.
4. Large scale data support capability. Although both softwares claim to be able to support 100000 level tasks, this capability is not blown out and needs to be verified in practice. On the one hand, control-m stores data in a database; on the other hand, it has actual cases (CCB); as a rookie, taskctl certainly does not have such big data cases. In addition, from the technical point of view, taskctl has no database, so it will encounter corresponding technical difficulties when it faces large-scale data support.
5. Centralized data management, software is always inseparable from data, scheduling software needs to manage a large number of process design information. As an enterprise platform, the centralized management of process information is necessary. Control-m stores data with data and centralized management; taskctl stores data in the form of files, which seems not to be centrally managed, and process information is stored on different scheduling service nodes.
6. Unified application portal, these two softwares can manage multiple scheduling servers at a single point, and different projects of enterprises can be managed and applied through unified client.
PK conclusion: 1From the perspective of enterprise characteristics, control-m has obvious advantages. Control-m is a real enterprise technology platform, and taskctl can only be called quasi enterprise technology platform at most. If control-m is a heavyweight scheduling platform, then taskctl can only be a lightweight scheduling platform.
Function point PK
Generally speaking, from the perspective of function, I think the two softwares are similar in terms of core scheduling function, application function and extension function. It’s just that the implementation is a little different. We take the core scheduling function as an example. The core function of scheduling is mainly determined by the judgment ability of task execution conditions. Control-m condition judgment is mainly determined by three aspects: resource condition, execution plan plan plan and user-defined condition; while taskctl is determined by resource condition, execution plan, structural condition (serial parallel structure, loop structure, etc.), fault tolerance condition, dependency, mutual exclusion, and user-defined condition. The common point of the two softwares is to expand and perfect the condition judgment system by defining the conditions, while the difference is that control-m is more abstract and taskctl is more specific.
If we have to talk about the difference of functions, I think that control-m has the function of file transfer (but this function is beyond the scope of scheduling), taskctl does not; taskctl has the function of process debugging, control-m does not.
PK conclusion: 1If only standing in the perspective of ETL scheduling and its application function point, the two softwares have their own merits, and PK results are equally divided.
User experience PK
When it comes to user experience, I don’t hesitate to vote for taskctl. The unique design of the software brings unique user experience, which is incomparable with control-m.
User experience is the core concept of software design. A software is not only a complete function, but also a friendly user experience. I remember that my former project leader emphasized user experience very much. Function is function and experience is experience. He often brings UI engineers and artists to discuss user experience. For a long time, I was deeply influenced by the leader. I think the focus of experience is UI, good art, good layout and good operation process. I think many friends agree with me. However, after contacting taskctl, my view has been greatly improved. I found that my understanding is too limited. Good experience is not only in the interface, but also from good architecture and mechanism. In order to have a good experience, I will not hesitate to innovate or even dare to break through. However, breakthrough and innovation have to pay a certain price, and experience innovation can not put the cart before the horse. As the official website of taskctl said, innovation is not the purpose, but better application is the root.
So, let’s look at how taskctl optimizes its user experience through a series of innovative designs.
Focus:The innovation of taskctl, key user scenarios, and comparison with control-m.
Let’s talk about several key innovations of taskctl
1. No data design, no data technology is not new, but in the field of professional dispatching technology platform, the software is the only one.
The core content of process design is to define the target tasks of various scheduling and control strategies of various tasks, such as dependency, parallelism, execution plan, etc. The traditional configuration method is used. The essence of this method is to store all kinds of designed information by designing various data tables, such as basic task information, control information, etc. when applying, the configuration mode is called configuration mode by designing various dialog boxes. Taskctl uses the development method to code the information of the process and develop the process like a development program. The process is designed through an integrated environment similar to vs.
3. Client offline application mode: no matter domestic professional scheduling software or foreign professional control-m, the application of client must be connected to the server; while taskctl client can be applied offline, that is, it does not need to connect to the server, which is to complete all operation experiences except real scheduling.
4. Plug in mechanism, professional scheduling platform to support different types of tasks is basic. Control-m is extended by line command, while taskctl explicitly proposes a driver plug-in mechanism to extend the support of different tasks through different driver plug-ins.
5. Various forms of application system, scheduling application of taskctl, not only has three graphical client software of admin, designer and monitor, but also has three matching payment client software. Both desktop client and background character interface client are complete application system. Although control-m has a background character interface, the application system is not complete and can not completely correspond to the foreground desktop client.
Key application scenarios
User experience is meaningful only when it comes to specific application scenarios,The most important application scenarios of scheduling include:
1. Installation and deployment application scenario, installation and deployment is the first scenario of software application.
2. Process design application scenario. For scheduling applications, this scenario may be the most important application scenario. Through this scenario, we tell the scheduling platform what to do and how to work.
3. Running the monitoring application scenario, needless to say, this scenario is the most concerned by customers, because we need to know how the scheduling platform is doing.
4. Query application scenarios, we are often very boring, always recall the past, to see what we have done.
Now, let’s take a look at how the innovation of taskctl can perform better than control-m in the above application scenarios.
1. Display effect of flow chart
Before analyzing each application scenario, let’s take a look at the display effect of flow chart. The quality of flow chart is related to many application scenarios.
The software is easy because it is easy to realize the specified business functions after mastering the technology. The difficulty of software is to realize some function, but it is not necessarily applicable. Whether it is a variety of familiar ETL tools scheduling, or many professional scheduling platform, have flow chart display. But if we say whose flowchart is more practical, I think taskctl’s is the most practical. Many software only stay at the level of drawing flow chart, and taskctl can not only draw flow chart, but also provide eight skills for beautiful and fresh display, convenient query, positioning, switching and other operations.
Although I said it for sure, different people have their own opinions. However, if you compare the eight features of taskctl one by one, you must remember that the fundamental purpose of a flow chart is not to draw a picture or design, but to show it intuitively and quickly understand what your process is like through graphics.
Although control-m graphic display has some skills, compared with taskctl, its skills seem to be much less; in addition, in front of large-scale graphs, taskctl wireless bars crisscross and regular display characteristics, which control-m can not cross.
2. Install and deploy application scenarios
Even if you are familiar with control-m, you can’t finish the environment construction for half a day. No matter whether you are familiar with taskctl or not, press the “taskctl novice experience” operation, 10 minutes to complete. Taskctl, whether it is a desktop client or a server, is almost a fool to install. The basic operation is: next, y, enter. On the one hand, the simplicity of taskctl installation is attributed to the simple design of the peripheral interface of the software and the design of the installation package itself, on the other hand, it is due to the design of no database.
3. Application scenarios of process design
In my opinion, the different applications in this scenario are the most different parts of taskctl. Generally speaking, both control-m and taskctl can realize the process design, but the control-m method lacks certain practical operability, while the taskctl method is not only convenient, but also simple and fast.
In a scheduling application, there are hundreds of tasks. If you define a thousand tasks through control-m, we will definitely Click Save switch back and forth in different dialog boxes, and each task may have many properties. It can be predicted that this kind of operation will make practical application more difficult. However, in practical application, when using control-m, many projects do not use the configuration mode provided by the software, but are defined through spreadsheets. Because the spreadsheet is a flat document after all, a lot of information can be edited in one place, so as to avoid many dialog box click switch operation. There is also an advantage of using a spreadsheet relative to a dialog box, that is, information search and location is much more convenient.
This phenomenon shows the following facts: when facing the application scenario of process design, in front of the large process, control-m has a complete implementation scheme in theory, but it is lack of operability in practice. The project would prefer to use the unrelated spreadsheet rather than the control-m’s own scheme, so that the control-m scheme is in vain.
Next, let’s talk about taskctl, which uses code to design the process. The code itself is carried by text, and a mature code integrated development and design environment is designed based on the code, which makes the design editing management of the process very convenient. For the concept of integrated development environment, we are very familiar with it. Graphics mode code mode can be arbitrarily switched, depending on personal preferences. Some people may think that the integrated development environment seems very good, but the code mode is easy to edit, but the learning cost of the code is high and the configuration is not easy to understand. Yes, it is. But unfortunately, the code of taskctl can only calculate the correct code. Although it has certain grammatical features, it is easy to understand and master. I can use it in less than half a day.
In addition, the process information scale with the same function designed by taskctl process code is the least, at least less than control-m. According to the official data of taskctl, the process information of taskctl is only 1 / 5 of that of control-m, or even less. I don’t think this figure is accurate. From the perspective of design, the control-m process information is not easy to calculate its scale. But I firmly believe that taskctl is the most concise, because it also has its own special mechanism and plug-in mechanism to ensure. As for how these mechanisms can ensure that the process information design is less and more concise, I won’t say much here. I’ll talk to you again when I have the opportunity.
4. Monitoring application scenarios
For this application scenario, in addition to some different operation skills, I don’t think taskctl has any outstanding highlights. But the complete background client application system, let the technical personnel have more choices.
5. Query application scenarios
For this scene, I think it is the most unconscious and absolutely intentional one in taskctl design. If you’re a technician, you’ll love it.
This surprise is due to the offline application mechanism of taskctl, which means that you can easily take your “process” around without relying on the server. It’s easy to know what your process looks like at any time. Go home, have a look, change; go to work during the day, whether it is the office, meeting room, break room, you are very convenient to discuss and discuss your process with colleagues; leave the project, you can take the process away quietly. One day, when you open the taskctl client, you can see the various processes you have designed. At that time, you must be very proud.
These seem to have nothing to do with scheduling, but they are not very practical?
How about control-m? My answer is, in theory, yes, but not in practice. You just need to think about whether the server is very convenient. Maybe in addition to the project site can be easily connected, other places, or wash sleep!
Finally, I want to say
Thank you so much for seeing this. PK to PK, conclusion to conclusion, choice to choose, everyone has their own choice, my choice is to face super large projects(more than 10000 tasks)ETL scheduling or control-m, while for small and medium-sized projects, I may choose taskctl.
Welcome to leave your comments in the comments area to discuss with usPriority will be given to 3-5 high-quality messagesGiveFree use of taskctl 6.0 permanent permissions, so that you can get a closer understanding of the functional attributes of the product, at the same time, it is also the support and affirmation of our domestic software R & D group