Home
Service Page : https://giipasp.azurewebsites.net/
Documentation :
- English : https://github.com/LowyShin/giip/wiki
- Japanses : https://github.com/LowyShin/giipdoc-ja/wiki
- Korean : https://github.com/LowyShin/giipdoc-ko/wiki
Samples
- Scripts : https://github.com/LowyShin/giip/tree/gh-pages/giipscripts
- giipAgent for UiPath
- giipAgent for Linux
- giipAgent for Windows
note: you may freely change or rebuild agent by giip agent source.
The CPU of the cloud server instance you are currently using is not usually displayed. So, would you like to find out on which hardware your server instance is running on with giip? -> Quick Start
-
The use of both AWS and Azure is increasing. Nowadays, when AI-dedicated machine prices are high, many are placing machines equipped with GPUs in offices and using them in conjunction with the cloud. In such cases, managing all resources with a single management system is impossible. However, try connecting all resources to giip! You can easily manage the current state!
-
When you need to check simple information, you often connect to the server. If it's one or two servers, it might be manageable, but what if it's a hundred? In this case, issue commands to one server with a simple command and have the results uploaded to KVS through simple scripting. Once you confirm that it has been uploaded successfully, go to the Script repository. The command you created a moment ago is saved there. Check and execute it on the remaining servers, and you're done!
-
There are often cases where CI/CD is vaguely cumbersome or doesn't work. In that case, if CI/CD tools themselves don't work, you eventually have to go into the server console, pull from git, compile, and upload. Jenkins is convenient, you say? With giip, you can just enter commands as if you were doing terminal work with the agent already present. Moreover, Jenkins is heavier than you might think, and it doesn't run at all on low-spec systems. Since giipAgent is a script-based agent, its CPU usage is not even 0.1%. Of course, if you upload heavy scripts, it will become heavy. If you upload heavy scripts to Jenkins, it adds to Jenkins' weight and can crash the system.
-
Are you using UiPath? UiPath is a very convenient RPA tool, but when you try to increase the number of tasks, significant costs arise. If you use giipAgent for UiPath, giip manages it centrally, regardless of UiPath's Orchestrator license. How about trying it once when the quantity is ambiguous?
What kind of design is needed to achieve fully automated operation of all devices in the world, including servers, PCs, mobile devices, IoT devices, etc.?
Most DevOps systems developed so far have been designed and created by developers, limiting the perspective to that of developers. Consequently, even if operational tools are created with DevOps, they may not cover vast systems, requiring manual intervention or forced sharding. What kind of system would result if the system architecture designed DevOps?
There exists an operational system architecture that has been continuously developed since 2007, based on experience in operating large-scale systems such as SFI (SONY Finance International)'s financial system, Nexon's large-scale online game system operation and its fight against hacking, as well as network management for services generating over 300Gbps of traffic in moments, like LoL (League of Legends), and the operation of national backbone-level ISP networks and large-scale cloud services.
That architecture is the giip engine.
Regardless of the tool used to create automation tasks, connecting to giip allows you to see how much your work has been reduced by the automation robot.
giip has been handling the workload of five people for the past month!
For an engine that has been continuously developed and updated for over a decade, the following philosophy is ingrained:
To collect data from all systems and use JSON as an unstructured data type database, the concept of Key Value Store (KVS) has been introduced. Historically, most system administrators have experienced the challenge of aligning data formats or switching tools due to mismatched data types. This challenge persists today. giip is the first operating system that presents data in a format directly viewable on the screen, providing intuitive management akin to entering servers, by displaying data in its original format, extracted from all operating systems such as Linux, Windows, MacOS, HP-UX, AIX, Solaris.
Designed based on Micro Service Architecture (MSA), where all functions can be separated and deleted, allowing easy integration of new technologies and ensuring that functional failures do not translate into service failures. Conversely, if only specific functions are needed from giip in other systems, they can be extracted and used independently. Additionally, giip's functionalities can be rearranged to create new services.
Similar to source code classification techniques like MVC or MVP, giip has its own unique MVC. While MVC in other development frameworks requires selecting all views, controllers, and models based on that framework, giip is designed to easily attach any model or view created using a completely different framework. It allows for straightforward utilization, such as purchasing and copying Bootstrap themes and customizing APIs, transforming giip into a new service. Whether using RDBMS or NoSQL, giip can be easily adapted.
Many developers today may consider the approach of managing through integrated source as outdated. However, companies like Amazon and GREE, operating more than 10,000 servers, deploy integrated source across numerous servers and activate only the required functionalities. This allows concentrating clients on servers among the 100, enabling scalability and redundancy by activating API functionalities on spare servers for different purposes and connecting them to GLB.
GLB binds numerous servers with a single function, distributing the load across resources worldwide, including servers distributed in the United States and Europe. It overcomes the limitations of L4, such as DSR, which prevents services beyond L4 performance. The concept of expansion and distribution, not possible in traditional L4, is applied.
Designed in 2007 based on global experience with operational automation systems such as HITACHI's JobNet, JP1, Senju, and BMC's Control-M. Unlike other solutions that retrofit features based on the term "RPA," giip's engine has a fundamental structure for true RPA. It significantly differs from products that forcefully added RPA features to existing solutions. To implement genuine user actions, various user input tools like Autohotkey, Selenium, Jmeter are utilized, allowing control and recognition of mouse movements and screen interactions similar to human behavior.
![alt text][giip-concept-AISE] [giip-concept-AISE]: https://github.com/LowyShin/giip/blob/gh-pages/images/PP/giip-PP-Intro-AISE02.png "giip-concept-AISE"
CQE is an engine that defines and manages tasks performed by clients, executing them based on specified conditions. This original giip engine is designed to manage tasks according to various work scenarios, such as checking anomalies on a specific server with a mobile device, issuing processing commands to a laptop, remotely operating the server, or transferring periodic backups over the network after local compression completes, deciding where to store them on media.
An exclusive message queue engine developed by giip to address the shortcomings of MQTT, exchanging data in JSON format. Regardless of the volume of tasks, it adheres to FIFO, enabling smooth asynchronous processing, much like using RDBMS cache, without entanglement. Authorized clients on the network (via Secret Key, etc.) can receive and assess this information or report on it. It is inherently equipped with KVS for easy creation of graphs, statistics, and features a parsing function for converting JSON data periodically into a format suitable for graphs, addressing format-related challenges in statistical analysis. Integration with graph tools based on JSON data, such as Morris.js and jqplot, makes it easier to create dashboards or reports.
Built on MSA architecture, giip can be seamlessly integrated with various open-source technologies like OpenStack, Kubernetes, Node.js, Bootstrap, ELK (Elastic Logstash Kibana), EFK (Elastic Fluentbit, Kibana). Users employing open-source tools can selectively use giip components, excluding redundant features. Conversely, it is possible to extract specific components from giip and use them in conjunction with other open-source solutions without making giip the main platform.
https://www.slideshare.net/LowyShin/giip-engine-open-source-revolution
You can sell various system operation know-how by scripting it. For example, if a simple backup script costs 10 won per run, running it once a day would be sufficient for a month, costing 300 won. Additionally, complex tools like MongoDB or Hadoop, difficult to install or configure, can be easily deployed on your server by selecting and deploying them from the marketplace. Now, you don't need to possess all the knowledge; just have the knowledge that helps your business, and for the rest, request it inexpensively from those with such experience. We will provide the most suitable system operation tools for global business.
Conversely, while system engineers used to be compensated only for their working hours, uploading their know-how can create infinite value.
Although machine learning features are currently utilized through CQE and MQE, we are evolving the marketplace by considering the structure connecting GPU resource providers, machine learning algorithm providers, and consumers of these resources. Of course, this is based on a thoroughly designed MSA, so it takes time, but even now, sensitive data like life insurance, app data, and stock-related data processing are being customized on the giip engine. The marketplace, based on this accumulated know-how, will continue to evolve, creating a world that connects resources, knowledge, and customers.
We have customers using AWS (Amazon Web Service). The VM and development servers, where the giip Agent is deployed, are in the local IP environment within the company. However, giip seamlessly manages all these environments. One day, we received a customer service call stating that they mistakenly applied iptables without assigning a remote access IP to an AWS VM, resulting in an inability to SSH from anywhere. AWS's response was that they had to create a new VM. Fortunately, the giip agent was operational, and a remote command service iptables stop (or systemctl stop firewalld on CentOS7) was executed. The server that was initially inaccessible suddenly became accessible as the iptables service was stopped.
Many challenges arise that cannot be addressed with just the basic tools provided by cloud service providers. Additionally, users often find themselves constrained by the rules set by these providers. Managing environments with local IPs, such as internal development servers, becomes difficult, and tools compatible with dynamic IP changes in domain-based services like Azure are scarce. Drawing from extensive operational experience in various environments, giip excels in managing such situations.
Traditional services like Cloud Service Brokerage (CSB) or Cloud Management Platform (CMP) merely connect to the APIs of individual cloud service providers. These services are often disregarded by those with actual system operation experience, as they are typically feature-rich but lack practicality. Most services focus on creating VMs, but the crucial aspect is ensuring that the services running on existing VMs or physical machines operate smoothly.
There are users who describe giip as having an irresistible charm that makes it hard to quit. Additionally, you can craft a brand-new UI entirely and manage it purely through APIs, almost like having your dedicated tool. Would you like to experience the allure of giip for yourself?
As long as you manage everything yourself, all services within the GIIP platform are free.
If you need paid services for storing a significant amount of log data, configuration support for features like backup and redundancy, and automation script assistance, you can opt for our paid services. Small-scale businesses can efficiently operate with cost-effective chatting support.
Choose the Reseller Plan if you want to manage your customers using GIIP. GIIP will handle billing, and we provide a revenue share. You can brand GIIP with your logo and use it as your dedicated site.
If you require support from dedicated experts, consider opting for Enterprise Support. Our infrastructure experts, with over 20 years of experience, will be dedicated to assisting with any issues in your infrastructure.
We are recruiting individuals or companies to join us in building giip! If you reach out to us, we will share the environment and source code, and we can work together to create giip.
Managing your systems, creating scripts, and operating them are all free. By standardizing your scripts and uploading them to the marketplace, you can even generate income. Are you still working on scripts alone without collaborating with others? How about joining forces to leverage your experience globally?
For all inquiries, collaborations, advertisements, and more, please contact us below.
Customers want to accomplish more with less budget, while workers seek fair compensation for their efforts. This tug-of-war can lead to legal troubles and dissatisfaction on one side. So, how can we bridge this gap?
GIIP Tokens address this issue by providing satisfaction to both parties. Customers specify tasks they deem within their budget. GIIP accepts the task if it aligns with the customer's budget. Workers perform the tasks for customers and receive compensation in GIIP Tokens.
GIIP Tokens can be sold on the aforementioned exchange, and giip utilizes part of the revenue to purchase GIIP Tokens.
Workers can sell GIIP Tokens at a lower price for quick cash or, if time permits, align their sales with giip's growth for greater profits.
This ecosystem is applicable to any IT outsourcing scenario. It allows for compromises between customer budgets and worker compensation preferences, enabling both parties to benefit financially.
If you need tasks done or want to become a worker, please contact us through the inquiry form!