{"id":74286,"date":"2020-03-15T20:42:50","date_gmt":"2020-03-15T17:42:50","guid":{"rendered":"https:\/\/prohoster.info\/blog\/administrirovanie\/monitoring-v-czode-kak-my-menyali-staruyu-bms-na-novuyu-chast-1"},"modified":"2020-03-15T20:42:50","modified_gmt":"2020-03-15T17:42:50","slug":"monitoring-v-czode-kak-my-menyali-staruyu-bms-na-novuyu-chast-1","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/administrirovanie\/monitoring-v-czode-kak-my-menyali-staruyu-bms-na-novuyu-chast-1","title":{"rendered":"Data Center Monitoring: How We Changed the Old BMS to a New One. Part 1","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

\"Data
\n<\/p>\n

What is BMS<\/h2>\n

\nA monitoring system for engineering systems in a data center is a key element of infrastructure, directly influencing a crucial metric for the data center, such as the response time of personnel to emergency situations and, consequently, the duration of uninterrupted operation.\u00a0<\/p>\n

BMS (Building Monitoring System) solutions are offered by many global vendors of equipment for data centers. During Linxdatacenter's operations in Russia, we have encountered various systems and faced diametrically opposed approaches from vendors regarding the operation of these systems.\u00a0<\/p>\n

We share how we completely updated our BMS system over the past year and why.\u00a0\u00a0
\n<\/a><\/noindex><\/p>\n

The root of the problem<\/h2>\n

\nEverything started 10 years ago with the launch of Linxdatacenter's data center in St. Petersburg. The BMS system, according to the industry standards of those years, was a physical server with installed software, accessed through a client program (the so-called 'thick' client).\u00a0<\/p>\n

There were few companies offering such solutions in the market at that time. Their products were the standard, the only answer to the existing demand. And credit should be given to them: both then and today, market leaders in general fulfill their basic task \u2013 providing functional solutions for data center operation.\u00a0<\/p>\n

A logical choice for us was the BMS solution from one of the largest global manufacturers. The chosen system at that time met all the requirements for monitoring a complex engineering facility, such as a data center.\u00a0<\/p>\n

However, over time, the requirements and expectations of users (that is, us, the data center operators) from IT solutions have changed. Large vendors, as a market analysis of the offered solutions has shown, were not prepared for this.<\/p>\n

The corporate IT market has been significantly influenced by the B2C sector. Digital solutions today must ensure the convenience of end-user work \u2013 that is the goal set by developers. This is evident from the improvement of user interfaces (UI) and the quality of user experience (UX) in many corporate applications.\u00a0<\/p>\n

People are getting used to comfort in all digital tools in their daily lives and expect the same from the tools they use for work tasks. They anticipate corporate applications to offer the same clarity, intuitiveness, simplicity, and transparency as the services available to them in financial services, taxi calls, or online shopping. IT specialists implementing solutions in the corporate environment also seek all the modern 'perks': easy deployment and scaling, fault tolerance, and unlimited customization options.\u00a0<\/p>\n

Large international vendors often overlook these trends. Relying on their long-standing authority in the industry, corporations tend to be rigid and inflexible when working with clients. The illusion of their own irreplaceability prevents them from seeing how young, tech-savvy companies are emerging right under their noses, offering tailored alternative solutions without the premium for brand name.<\/p>\n

Drawbacks of the old BMS system\u00a0<\/h2>\n

\nThe most significant downside of our outdated BMS solution was its slow performance. Investigating several incidents related to the response delays of on-duty personnel helped us understand that events were sometimes displayed in the BMS system with significant delays. The system was neither overloaded nor malfunctioning; it was simply that the versions of its components (e.g., JAVA) were outdated and could not work properly with newer operating system versions without updates. These updates could only be performed along with the BMS system itself, and the vendor did not ensure automatic version succession, making the process almost as labor-intensive for us as switching to a new system, while the new solution retained some of the shortcomings of the old one.\u00a0\u00a0<\/p>\n

Let's add a few more unpleasant 'details':<\/p>\n

    \n
  1. Charges for connecting new devices on the principle of 'one IP address \u2013 one paid license';\u00a0\n<\/li>\n
  2. The inability to update software without purchasing a support package (referring to the upgrade of free components and fixing bugs in the BMS program itself);\n<\/li>\n
  3. High cost of support;\u00a0\n<\/li>\n
  4. Operating on a 'bare metal' server, which can fail and has limited computing resources;\n<\/li>\n
  5. \u2018Redundancy\u2019 by installing a second bare metal server with a duplicating license package. At the same time, synchronization of databases between the main and backup servers is absent \u2013 meaning manual database transfer and a long switch to the backup;\n<\/li>\n
  6. A 'thick' client for users, not accessible externally, without mobile device extension and remote access options;\n<\/li>\n
  7. A stripped-down web interface without graphics cards and sound notifications, accessible externally but practically unused by employees due to its lack of informativeness;\n<\/li>\n
  8. Absence of animation in the interface \u2013 all graphics consist only of a 'background' image and static icons. As a result, the overall level of visibility is low;\n

    Everything looked roughly like this:<\/p>\n

    \"Data
    \n
    \n\"Data<\/li>\n

  9. Limitation in creating virtual sensors \u2013 only the addition function is available, whereas real sensor models require the ability to perform a set of mathematical operations for accurate calculations reflecting the realities of operations;\u00a0\n<\/li>\n
  10. Inability to obtain real-time data or archived data for any purposes (for example, for display in the client's personal account);\n<\/li>\n
  11. Complete lack of flexibility and ability to modify the BMS for existing data center processes.\u00a0\n<\/li>\n<\/ol>\n

    Requirements for the new BMS system<\/h2>\n

    \nConsidering the above, our main requirements are as follows:<\/p>\n

      \n
    1. Two independent mutually redundant machines with automatic synchronization, operating on two different cloud platforms in different data centers (in our case, the Linxdatacenter data centers in St. Petersburg and Moscow);\n<\/li>\n
    2. Free addition of new devices;\n<\/li>\n
    3. Free software and component updates (excluding functional improvements);\n<\/li>\n
    4. Open source code, allowing us to maintain the system ourselves in case of issues on the developer's side;\n<\/li>\n
    5. Ability to obtain and use data from the BMS, for example, on the website or in the personal account;\n<\/li>\n
    6. Access via a web browser without a 'thick' client;\n<\/li>\n
    7. Use of domain accounts for employees to access the BMS;\n<\/li>\n
    8. Presence of animation and many other small and not-so-small wishes that materialized into a detailed specifications document.\n<\/li>\n<\/ol>\n

      The Last Straw<\/h2>\n

      \n\"Data
      \n
      \nAt the moment we realized that the data center had outgrown its BMS, the most obvious solution seemed to be upgrading the existing system. \"Never change horses in midstream,\" right?\u00a0<\/p>\n

      However, large corporations typically do not offer custom modifications to their decades-old, \"polished\" solutions sold in dozens of countries. While young companies focus on testing ideas or prototypes of future products with potential customers and rely on user feedback for product development, corporations continue to sell licenses for what was once a truly great product that, unfortunately, is now outdated and inflexible.<\/p>\n

      And we felt the difference in approach ourselves. During correspondence with the manufacturer of the old BMS, it quickly became clear that the vendor's proposed upgrade to the existing system would essentially turn into a purchase of a new system with a semi-automated database transfer, high costs, and pitfalls during the transfer that even the manufacturer couldn't predict. Of course, this increased the cost of support for the upgraded solution, and there remained the need to purchase licenses for any expansion.<\/p>\n

      And the most unpleasant part \u2013 the new system could not fully meet our requirements for redundancy. The upgraded BMS could be implemented as we wanted, on a cloud platform, which would allow us to abandon the \"hardware\", but the backup option was not included in the cost. For data redundancy, we would have to buy a second virtual server<\/a> BMS and an additional set of licenses. With the cost of one license being around $76 and the quantity (the key to connect to is specified, and iroh finds the associated host and establishes an encrypted connection using the QUIC protocol). Direct P2P connections are established whenever possible, but if not, it falls back to using relays, which are also employed for host discovery by keys. You can run your own relay or connect to public relays supported by the community.<\/a> of 1000 units, this adds up to an additional $76,000 in expenses just for licenses for the backup machine.\u00a0<\/p>\n

      The \"cherry on top\" in the new version of BMS became the requirement to purchase additional licenses \"for all devices\" \u2013 even for the main server. It should be clarified that there are devices connected to BMS via gateways. A gateway has one IP address but controls several devices (on average 10). In the old BMS, one license was required per IP address of the gateway, and the statistics looked approximately like this: \"IP addresses\/licenses 1000, devices 1200.\" The updated BMS operated on a different principle, and the statistics would look like this: \"IP addresses 1000, devices\/licenses 1200.\" This means that the vendor changed the licensing assignment principle in the new version, and we had to purchase about 200 additional licenses.\u00a0<\/p>\n

      The budget for the \"upgrade\" ultimately consisted of four items:\u00a0<\/p>\n