Interview of TAdviser with the chairman of the board of Yugra Bank Dmitry Shilyaev
Technical failure in work of Yugra Bank forced to worry enough investors: within two days there were problems with accomplishment of banking activities. What became the reason of this incident, who is guilty and what measures are taken that similar did not repeat, in an interview of TAdviser told the chairman of the board of Yugra Bank Dmitry Shilyaev.
Shilyaev the Main output - needs to be carried to
Failure which occurred at the end of April, it something extraordinary for bank? Whether there were such incidents earlier?
Dmitry Shilyaev: No, nothing similar was before. Failure arose because two fundamental nodes - uninterruptible power supply unit and a storage system in a row refused. At violation in electricity feed-in the uninterruptible power supply unit did not switch to batteries, and the storage system lost the configuration.
DWH by determination, on the characteristics stated by vendor, is non-volatile. At power off it, according to the description, has sufficient potential for reset of data from a cache, including a configuration, on the hard drive. Why it did not occur, there is no answer yet. As well as on why the uninterruptible power supply did not switch to the battery. Both a system, and the uninterruptible power supply at us from one supplier – IBM.
And what happened because of a hardware failure?
Dmitry Shilyaev: We lost access to DWH, an important part of the main infrastructure. The threat of data loss of clients at the same time was not, however DWH could not function, respectively, the virtual machines installed on it did not work. Once again I want to note, data on clients were not mentioned.
Recovery of servers using which we provided data access borrowed in the afternoon, from 1:42 p.m. on April 25 when there was a failure, up to 2 o'clock in the morning on April 26. Recovery of DWH took more time, it was performed by engineers of IBM. According to their pro forma statement, there was "an interruption of working capacity after failure in network of power supply".
How many days lasted problems?
Dmitry Shilyaev: Complete interruption of working capacity lasted 12 hours approximately. Then the data access was recovered, but the being available server resources were not in that volume which would allow to be connected to a system to all operators therefore offices worked on April 26 in the limited mode.
On Thursday already 95% of all power functioned April 27, in general we provided a full service of all clients.
What conclusions were drawn from the incident?
Dmitry Shilyaev: The main output - needs to be carried territorially data processing centers. The main and reserve DPCs were placed with us in one territory. DPCs of own construction, we only rent the premises. Complaints to them were not before.
Now we build the third additional DPC on other platform.
And there were ideas to use outsourcing DPC?
Dmitry Shilyaev: We considered a question of outsourcing, lease of racks, but did not make such decision - risks too serious.
It is necessary to make additional DPCs and to carry them territorially. A few years ago we even considered option of placement of DPCs in the different cities with periodic next switching of work to each of them. But did not implement the project because it was quite costly. Perhaps, we will return to this discussion, but later. Now stopped that in other territory in Moscow we do one more reserve DPC.
Requirements of the Central Bank regulate at what distance there have to be reserve DPCs?
Dmitry Shilyaev: No, there are only requirements about need of reservation. At the time of technical failure of the Central Bank showed concern too, there came employees who participated in our meetings.
Moscow - same expensive city?
Dmitry Shilyaev: It is clear, that in Orenburg or, for example, Obninsk will build DPC cheaper, but the main question remains in the personnel. As the bank rather dynamically develops, the DPC should develop after bank - to increase the capacities, to pass to the new platform, etc., and all this requires attention of personnel and its development. This question forces to concentrate on Moscow.
How do you estimate actions of IT bank staff at the time of critical situation?
Dmitry Shilyaev: In general it is professional. There are no claims from IBM, notes on hardware maintenance and updates. Moreover, we updated a microcode literally some days before the event. A similar situation on the uninterruptible power supply – we caused service company, it also did not have claims to operation - the equipment at us in ideal conditions now. All equipment new, was purchased in the middle of 2015 and still is on a guarantee.
In new DPC you are going to use IBM too? Or, maybe, for example, HP for secure?
Dmitry Shilyaev: To HP there are requirements too. I personally consider that IBM is after all other level. Usually HP is a core banking system, IBM is a processing because processing is 24/7 with the minimum time of possible idle time and minimum holding time. Besides we are strongly tied on IBM: on this equipment there are competences, and our infrastructure is strongly connected. I mean the systems of synchronization of DWH, the system of virtualization. At us not only x86, but also Power processors with the AIX operating system, the built-in systems of virtualization and so on. Integration is rather close.
The bank has an IT strategy? What its priorities?
Dmitry Shilyaev: Of course, IT strategy exists, in its basis - support of development of retail business, increase in fault tolerance. Support of retail is, first of all, remote channels of bank service, development of Internet bank, card business. And also development of the website of bank, mobile applications and set of other services.
And in general, how you see further development of banking sector?
Dmitry Shilyaev: We are a few conservatives in this sense. There are fintekh, there are dynamic startups, but the bank is, first of all, reliability. The bank undertakes others risks, and in this paradigm it should provide the reliable back and the base.
Now there is a certain trend - the open interface. The bank is some kind of safe to which interface can build infinite number of the financial technical-companies. Peculiar "middle office" - what calculates risks and determines the cost of risks. Actually this approach allows bank not to spend time and resources for guessing of momentary desires of clients and to perform the basic function, and on the other hand – to be present practically everywhere. The second trend is a marketplace. Using the network of offices, the bank can help to sell adjacent products.
| The site content is translated by machine translation software powered by PROMT. The machine-translated articles are not always perfect and may contain errors in vocabulary, syntax or grammar. Read original article If you find inaccuracies or errors in the results of machine translation, please write to editor@tadviser.ru. We will make every effort to correct them as soon as possible. |