“The biggest competitor is those who do everything themselves” - Petr Zaitsev about the creation and development of the Percona company
We continue to talk with Percona specialists. The material was very interesting. Our questions were answered by Peter Zaitsev , the founder and CEO of the company. Peter's long record is great: the creation and development of the largest consulting, engaged in support and maintenance of solutions based on MySQL and MongoDB; co-author of the book MySQL. Performance Optimization; Regular publications in the Percona Database Performance Blog, one of the best technical blogs on MySQL and PostgreSQL. Shortly before the founding of his own company, Peter headed the performance optimization group (High Performance Group) in MySQL AB.
We talked with Peter about the history of the company's formation, key values, the difficulties of working on a modern technology market and the company's development strategy, discussed some features of the selection of employees in Percona - a process that Peter oversees personally.
For the first time at PG Day'17 Russia, Peter will read his author’s course on InnoDB architecture and performance optimization for the Russian-speaking audience, as well as review available tools for using MySQL 5.7 as a document-oriented database .
')
PG Day: Peter, tell me about yourself, who are you and what are you doing?
Peter: Peter Zaitsev - founder and CEO of the company “Perkona” (Percona) . The company was founded more than 10 years ago. We provide solutions in the field of MySQL and MongoDB. Our main business is support (support), remote administration (remote DBA) and management (managed services), but we also deal with consulting and training.
PG Day: How did you start building a business around MySQL? Why did you choose this DBMS?
Peter: In fact, it was not like this that I sat down and began to think on what to do business. Before the founding of Perkona, I worked at MySQL AB and, accordingly, I knew the technology, I had connections, and I met people in this industry. And so, when the opportunity and desire to make something of their own appeared, the choice was quite clear and obvious. In business, you need to do what you can, in what you see potential success. At the time of founding “Perkony” I didn’t have a grand idea that it would become a company in which more than 100 people would work. I thought that I would do consulting for a couple of years, and then maybe I would do something different. But the beginning turned out to be quite successful, everything went and went, and we got involved.
PG Day: When you were still working in MySQL AB, did you have a technical background, so you came to technology management?
Peter: Yes, that's right. In MySQL, I worked in a purely technical role. Last year I was the entry-level manager. But, basically, it was a technical direction.
PG Day: If you were to decide now to open a consulting business, by what criteria would you choose a technology stack? How would your decision regarding current realities change?
Peter: There are several criteria here. I believe that with MySQL I was very lucky for two reasons. First, when Percon was founded, the MySQL market was growing very fast , there were not enough specialists, and many companies that wanted to introduce this technology were ready to pay a lot of money. Such cycles are repeated for other technologies. For example, Hadoop. Several years ago, when the explosive growth of Hadoop occurred, people who understood this technology had very good opportunities, because demand exceeded supply by several times . Now, of course, I would look at this direction. And second, one of the key moments of success - it is important to do what you like, what you have a soul for . If you do what you don’t like, just for the sake of money, then there’s hardly any success, no matter what field you’re working with.
PG Day: All your software is completely free. You do not make proprietary software that could be sold. Why did you refuse such a business development model as selling your closed decisions?
Peter: The market for proprietary software or even open source core functionality (open core) with some paid features is much larger than the market for fully open software that Percona is involved in. However, it also has a lot of competing companies. A lot of resources are invested in this market, companies are raising a lot of money - venture capital, etc. They have a completely different business model, which is not very interesting to us.
We focus on the market of clean, open source software , which allows us to seriously differ from other vendors. This is especially important in the MySQL market, where we do not work with any completely new technology, we do not create it from scratch.
Often people, speaking of the Percons market, compare us with someone like MongoDB, MySQL or MariaDB. In fact, this is not entirely correct, because our task is not to create a platform or technology from scratch, but to provide services for the best open platform in this area. If we did Percona Server with closed lotions, the difference between our solution and, for example, Oracle solutions would be very small. It would be the same Oracle, but simpler, cheaper. This is not a very good motivator. If Oracle is a proprietary software that makes users dependent on the vendor (vendor lock-in) , even in the case of MySQL it is an open core, then we make Open Source, which helps to avoid this. Very clear and simple difference.
PG Day: How from the moment your company was founded, did the support market develop, consulting MySQL? Surely the "Perkony" has more competitors. How are you different from them? What allows you to remain competitive and hold high positions in the market?
Peter: Speaking of MySQL, Perkona operates in several markets that have different competitors. If we talk about support , companies that are our competitors are only Oracle and MongoDB , the only two companies that can actually provide a full cycle of support, including fixing problems, bugs in software. MariaDB was also in the MySQL market, but now they are coming out of it and are focusing only on supporting the MariaDB platform.
We are unique in that we are the only company that supports a wide range of MySQL and MongoDB technologies , and not just our solutions. For example, if MySQL from Oracle supports its technology stack for backup, monitoring, and everything else, then we support both our Percona Server technology, Percona XtraDB Cluster, and MySQL from Oracle, MySQL option from MariaDB, and cloud solutions such as Amazon RDS and Google Cloud SQL.
In the area of support, we are the only vendor that covers both MySQL and MongoDB at the same time, which is interesting enough for many people for two reasons. First, when technology evolves, many companies have a balance between how much MySQL and how much MongoDB they use. It's hard enough to work with two vendors, it is inflexible and uncomfortable. If the price of Oracle services “takes off”, you will say: “Listen, guys. We bought support for 100 MySQL instances from you, but in fact we will use 50 MySQL instances, and the rest will be MongoDB. Can we transfer licenses? ”. Of course, they will say: “You cannot”, because they will receive less money. For our part, we provide support that covers both solutions, and how much the client expects to use is left to his discretion.
Secondly, any vendor that works with only one technology is interested in using this technology as much as possible. The more you can show that MySQL can be used, the more support contracts you can get. At the current stage, if there is a choice between these two DBMS, we do not have such a motivation. The client can use MySQL or MongoDB - for us it does not make a financial difference. This allows our entire team to offer more objective solutions that are based on customer benefits and avoid conflicts of interest. Probably one of the most important things that formed the basis of “Perkona”: we try to concentrate on a solution that is really the best customer, and not for our wallet.
In the field of remote management and administration, our competitors are already different: companies such as Pythian and Datavail. And also “cloud” companies, because the owners of many businesses are now beginning to think: “If I use Database as a Service, cloud solutions from Amazon, Google Cloud SQL, then maybe I won't need DBA?” For many companies This is not entirely true, but this option of competition is also an interesting case for us. But if we talk in general about our services, the biggest competitor is probably those who do everything on their own . I often explain this by the example of the office cleaning market. Who is the biggest competitor here? Perhaps, many will say: “But let us save money, we will clean and vacuum ourselves!” It is clear that this is a kind of choice. And this choice is faced by people who decide to manage their own databases or give them to someone else.
PG Day: You mentioned the remote administration service. As I understand it, these are DBA specialists who work as external contractors for companies that provide consulting services. Can you tell us how you came to this direction, than it is attractive now?
Peter: First, we prefer to talk about this not as an external contract, which is often associated with something negative. We often talk about the continuation of the cloud model and technologies like “X as a service” . What is going on with the cloud? People realized that very often it does not make sense to hire and maintain a full staff of specialists who can install servers and connect cables. For many companies, these are not very critical decisions; they can be outsourced so that the company can focus on its core business.
For tech companies, this is product creation. For companies in the field of retail sales it can be marketing or logistics. And database maintenance operations are often simply reduced to work that needs to be done. If they try to do it inside the company, then it takes energy. In this case, we offer people to convey all these problems to us and focus on developing their product .
In this regard, the Database as a Service market even helps us, because these companies invest a lot in marketing. They try to convey to customers the idea that it may not be profitable to set up databases themselves and manage them. But they do not go to the end. If you look, for example, on Amazon RDS, then this is not the service that the manager can work with. Someone has to decide which instances need to be deployed, how to design a scheme. Simple setup of backup, monitoring also requires the presence of technical specialists. And this is exactly what we can offer, because we are able to become part of a team and cooperate directly with the management and the development department in order to completely solve these problems.
PG Day: You told the diversification of the technologies that you support. At what point did the decision to take on MongoDB support appear? Was it a sensible step, or was it so organically?
Peter: On the one hand, we wanted to add more technology at the beginning of the development of our strategy. I consider it very important that we really can provide objective solutions to clients. There is a saying: if you have a hammer, everything looks like a nail - if you have one tool that you can work with, then everything seems to be suitable for this tool, even when it is not needed. There was an opportunity in the MongoDB market, and it came very well for us: there is a demand for alternative support vendors in this market, in part because MongoDB, as a company, has worked a lot to position itself as an alternative to MySQL.
We were able to acquire the company TokuTek , which already provided support services to MongoDB and its own “fork” MongoDB, known in the market. This allowed us to successfully enter a new market, not to do everything from scratch. If this did not happen, we would need more time to start doing something other than MySQL.
PG Day: Consulting is a business that is based on people. The company you have is distributed, there is not one big building in which you are all sitting. How do you manage to cope with a large number of clients? What specialists do you have?
Peter: We currently employ about 150 people, 30 different countries, 20 different US states. Mostly, people work from home. In the United States, in North Carolina, we have a small office, with an average of 5 people per day. It is made so that it was convenient to hold meetings there. Guests often come to us. We also have many small offices.
In my opinion, it is not entirely correct to put the support business and the consulting business in one row. We started as a consulting company , and after a few years it became difficult for us to develop. The company grew, its structure became more complex , its costs increased, we began to develop software such as Percona Server for MySQL . Consulting is often quite unpredictable: there are periods when clients come one after another, but sometimes the less “fruitful” season happens, and then you have a lot of consultants who have nothing to do.
Therefore, we switched to a business model: support and remote management is a fixed-price subscription for a month or a year, which gives an advantage to both us and our customers. For us, this is a predictable income , the ability to plan investments and resources. But for customers it is also more predictable on budgets that can be planned once. Over the past year, we switched to this subscription service model, and somewhere around 85% of our revenues are support and remote management . And only 15% is consulting and training.
Now in terms of scaling . Having a fixed subscription allows us to invest in technology in order to increase our efficiency. This is most obvious in terms of remote control. Many companies that charge by the hour have a conflict of interest. If you introduce too much automation to customers, they are required to pay less hours. But since customers pay you for hours, it can be difficult to find money to invest in innovation in order to increase automation and increase productivity. From our point of view, since we sell the result for a fixed price, the main interest is to automate everything as much as possible in order to reduce human costs. This allows us to seriously invest in automation . This is beneficial to the client, because systems that are well automated are more stable. The computer makes fewer mistakes than people. This is especially true of large systems and large teams.
I think a large proportion of our success is associated with this. This allows us to be much more flexible in terms of staff. Suppose you sell some billing to a client, and you have some super-developer in your staff that is five times more productive, and therefore you would like to take three times more money for it. In fact, the consulting business model does not work that way. Corporate customers are negatively disposed to a large range in prices, although it is widely known that productivity in the field of software development between middle and high-end specialists may differ tenfold. Therefore, focusing on our subscription model, we can afford to invest in highly efficient personnel .
PG Day: You started talking about corporate-level customers. I think it’s no secret that you have a lot of very large clients like BBC, Airbnb, Cisco and so on. What are the main difficulties of interaction with a client of this level?
Peter: Large companies have higher requirements for structure and consistent results . This was for me one of the challenges in the transformation “Perkony”. When we started, we worked with many startups. Interesting, active guys who could say "Listen, how cool: I ask Perkone a question and get three different opinions from them that have a right to exist." We all understand that in the field of technology there is no one correct solution, as in the problem of mathematics. Usually there are several, and all are quite acceptable: you can do this or that, and everything will work. At first we hired such smart, independent guys and said: “Guys, do what you want. The main thing is that it is reasonable, and the client was satisfied. "
When we started working with enterprise clients, expectations began to change . They want consistency - like, for example, in Starbucks, who say: “Guys, we don’t need the best coffee in the world, but we want to come in, ask for a cappuccino, and may it always be absolutely the same.” Corporations pay a lot of attention to this aspect. They want to ask the same question (which they often do because they have many teams that are facing the same problems) and expect to receive the same answer. If they get a different answer each time, it just drives them crazy. And this requires a completely different approach to team organization . , , , , “” . 25 , , . , . enterprise- .
PG Day: ? , . , , .
: , , — . : , , . . - -, , . , . , . — , . : «, SQL . ». , - , MySQL , .
? , . , , . , , . . : , . , . , , : « ?» , , , . “”, , , : “” , . , , , . .
95% . , , . , , , , , , , , , . Joke, of course. , , , . , .
. : , . , , , 5 , , , . , , . -, , . , “”, : «, . , , , !» , . , , . -, , , . .
PG Day: ? DBA — , “” , - .
: , , . , - , . , , , , . , . , background check : - , , -. . , 10 . - , - . , , , - — .
PG Day: , , , -, PG Day'17.
: -, InnoDB. , storage engine, , , 99% MySQL . Percona Live , . : MySQL, - . , . . - , InnoDB: , , , , . , , InnoDB , , “” - , .
PG Day: Who can be interested in participating in a master class in the first place?
Peter: Most of all this master class is interesting DBA and system administrators who work with MySQL. They need to know how to use the maximum MySQL capabilities. Listeners will have many chances to ask questions, discuss some of their applied problems in the context of the material being studied.
PG Day: Thank you, Peter!
We talked with Peter about the history of the company's formation, key values, the difficulties of working on a modern technology market and the company's development strategy, discussed some features of the selection of employees in Percona - a process that Peter oversees personally.
For the first time at PG Day'17 Russia, Peter will read his author’s course on InnoDB architecture and performance optimization for the Russian-speaking audience, as well as review available tools for using MySQL 5.7 as a document-oriented database .
')
PG Day: Peter, tell me about yourself, who are you and what are you doing?
Peter: Peter Zaitsev - founder and CEO of the company “Perkona” (Percona) . The company was founded more than 10 years ago. We provide solutions in the field of MySQL and MongoDB. Our main business is support (support), remote administration (remote DBA) and management (managed services), but we also deal with consulting and training.
PG Day: How did you start building a business around MySQL? Why did you choose this DBMS?
Peter: In fact, it was not like this that I sat down and began to think on what to do business. Before the founding of Perkona, I worked at MySQL AB and, accordingly, I knew the technology, I had connections, and I met people in this industry. And so, when the opportunity and desire to make something of their own appeared, the choice was quite clear and obvious. In business, you need to do what you can, in what you see potential success. At the time of founding “Perkony” I didn’t have a grand idea that it would become a company in which more than 100 people would work. I thought that I would do consulting for a couple of years, and then maybe I would do something different. But the beginning turned out to be quite successful, everything went and went, and we got involved.
PG Day: When you were still working in MySQL AB, did you have a technical background, so you came to technology management?
Peter: Yes, that's right. In MySQL, I worked in a purely technical role. Last year I was the entry-level manager. But, basically, it was a technical direction.
PG Day: If you were to decide now to open a consulting business, by what criteria would you choose a technology stack? How would your decision regarding current realities change?
Peter: There are several criteria here. I believe that with MySQL I was very lucky for two reasons. First, when Percon was founded, the MySQL market was growing very fast , there were not enough specialists, and many companies that wanted to introduce this technology were ready to pay a lot of money. Such cycles are repeated for other technologies. For example, Hadoop. Several years ago, when the explosive growth of Hadoop occurred, people who understood this technology had very good opportunities, because demand exceeded supply by several times . Now, of course, I would look at this direction. And second, one of the key moments of success - it is important to do what you like, what you have a soul for . If you do what you don’t like, just for the sake of money, then there’s hardly any success, no matter what field you’re working with.
PG Day: All your software is completely free. You do not make proprietary software that could be sold. Why did you refuse such a business development model as selling your closed decisions?
Peter: The market for proprietary software or even open source core functionality (open core) with some paid features is much larger than the market for fully open software that Percona is involved in. However, it also has a lot of competing companies. A lot of resources are invested in this market, companies are raising a lot of money - venture capital, etc. They have a completely different business model, which is not very interesting to us.
We focus on the market of clean, open source software , which allows us to seriously differ from other vendors. This is especially important in the MySQL market, where we do not work with any completely new technology, we do not create it from scratch.
Often people, speaking of the Percons market, compare us with someone like MongoDB, MySQL or MariaDB. In fact, this is not entirely correct, because our task is not to create a platform or technology from scratch, but to provide services for the best open platform in this area. If we did Percona Server with closed lotions, the difference between our solution and, for example, Oracle solutions would be very small. It would be the same Oracle, but simpler, cheaper. This is not a very good motivator. If Oracle is a proprietary software that makes users dependent on the vendor (vendor lock-in) , even in the case of MySQL it is an open core, then we make Open Source, which helps to avoid this. Very clear and simple difference.
PG Day: How from the moment your company was founded, did the support market develop, consulting MySQL? Surely the "Perkony" has more competitors. How are you different from them? What allows you to remain competitive and hold high positions in the market?
Peter: Speaking of MySQL, Perkona operates in several markets that have different competitors. If we talk about support , companies that are our competitors are only Oracle and MongoDB , the only two companies that can actually provide a full cycle of support, including fixing problems, bugs in software. MariaDB was also in the MySQL market, but now they are coming out of it and are focusing only on supporting the MariaDB platform.
We are unique in that we are the only company that supports a wide range of MySQL and MongoDB technologies , and not just our solutions. For example, if MySQL from Oracle supports its technology stack for backup, monitoring, and everything else, then we support both our Percona Server technology, Percona XtraDB Cluster, and MySQL from Oracle, MySQL option from MariaDB, and cloud solutions such as Amazon RDS and Google Cloud SQL.
In the area of support, we are the only vendor that covers both MySQL and MongoDB at the same time, which is interesting enough for many people for two reasons. First, when technology evolves, many companies have a balance between how much MySQL and how much MongoDB they use. It's hard enough to work with two vendors, it is inflexible and uncomfortable. If the price of Oracle services “takes off”, you will say: “Listen, guys. We bought support for 100 MySQL instances from you, but in fact we will use 50 MySQL instances, and the rest will be MongoDB. Can we transfer licenses? ”. Of course, they will say: “You cannot”, because they will receive less money. For our part, we provide support that covers both solutions, and how much the client expects to use is left to his discretion.
Secondly, any vendor that works with only one technology is interested in using this technology as much as possible. The more you can show that MySQL can be used, the more support contracts you can get. At the current stage, if there is a choice between these two DBMS, we do not have such a motivation. The client can use MySQL or MongoDB - for us it does not make a financial difference. This allows our entire team to offer more objective solutions that are based on customer benefits and avoid conflicts of interest. Probably one of the most important things that formed the basis of “Perkona”: we try to concentrate on a solution that is really the best customer, and not for our wallet.
In the field of remote management and administration, our competitors are already different: companies such as Pythian and Datavail. And also “cloud” companies, because the owners of many businesses are now beginning to think: “If I use Database as a Service, cloud solutions from Amazon, Google Cloud SQL, then maybe I won't need DBA?” For many companies This is not entirely true, but this option of competition is also an interesting case for us. But if we talk in general about our services, the biggest competitor is probably those who do everything on their own . I often explain this by the example of the office cleaning market. Who is the biggest competitor here? Perhaps, many will say: “But let us save money, we will clean and vacuum ourselves!” It is clear that this is a kind of choice. And this choice is faced by people who decide to manage their own databases or give them to someone else.
PG Day: You mentioned the remote administration service. As I understand it, these are DBA specialists who work as external contractors for companies that provide consulting services. Can you tell us how you came to this direction, than it is attractive now?
Peter: First, we prefer to talk about this not as an external contract, which is often associated with something negative. We often talk about the continuation of the cloud model and technologies like “X as a service” . What is going on with the cloud? People realized that very often it does not make sense to hire and maintain a full staff of specialists who can install servers and connect cables. For many companies, these are not very critical decisions; they can be outsourced so that the company can focus on its core business.
For tech companies, this is product creation. For companies in the field of retail sales it can be marketing or logistics. And database maintenance operations are often simply reduced to work that needs to be done. If they try to do it inside the company, then it takes energy. In this case, we offer people to convey all these problems to us and focus on developing their product .
In this regard, the Database as a Service market even helps us, because these companies invest a lot in marketing. They try to convey to customers the idea that it may not be profitable to set up databases themselves and manage them. But they do not go to the end. If you look, for example, on Amazon RDS, then this is not the service that the manager can work with. Someone has to decide which instances need to be deployed, how to design a scheme. Simple setup of backup, monitoring also requires the presence of technical specialists. And this is exactly what we can offer, because we are able to become part of a team and cooperate directly with the management and the development department in order to completely solve these problems.
PG Day: You told the diversification of the technologies that you support. At what point did the decision to take on MongoDB support appear? Was it a sensible step, or was it so organically?
Peter: On the one hand, we wanted to add more technology at the beginning of the development of our strategy. I consider it very important that we really can provide objective solutions to clients. There is a saying: if you have a hammer, everything looks like a nail - if you have one tool that you can work with, then everything seems to be suitable for this tool, even when it is not needed. There was an opportunity in the MongoDB market, and it came very well for us: there is a demand for alternative support vendors in this market, in part because MongoDB, as a company, has worked a lot to position itself as an alternative to MySQL.
We were able to acquire the company TokuTek , which already provided support services to MongoDB and its own “fork” MongoDB, known in the market. This allowed us to successfully enter a new market, not to do everything from scratch. If this did not happen, we would need more time to start doing something other than MySQL.
PG Day: Consulting is a business that is based on people. The company you have is distributed, there is not one big building in which you are all sitting. How do you manage to cope with a large number of clients? What specialists do you have?
Peter: We currently employ about 150 people, 30 different countries, 20 different US states. Mostly, people work from home. In the United States, in North Carolina, we have a small office, with an average of 5 people per day. It is made so that it was convenient to hold meetings there. Guests often come to us. We also have many small offices.
In my opinion, it is not entirely correct to put the support business and the consulting business in one row. We started as a consulting company , and after a few years it became difficult for us to develop. The company grew, its structure became more complex , its costs increased, we began to develop software such as Percona Server for MySQL . Consulting is often quite unpredictable: there are periods when clients come one after another, but sometimes the less “fruitful” season happens, and then you have a lot of consultants who have nothing to do.
Therefore, we switched to a business model: support and remote management is a fixed-price subscription for a month or a year, which gives an advantage to both us and our customers. For us, this is a predictable income , the ability to plan investments and resources. But for customers it is also more predictable on budgets that can be planned once. Over the past year, we switched to this subscription service model, and somewhere around 85% of our revenues are support and remote management . And only 15% is consulting and training.
Now in terms of scaling . Having a fixed subscription allows us to invest in technology in order to increase our efficiency. This is most obvious in terms of remote control. Many companies that charge by the hour have a conflict of interest. If you introduce too much automation to customers, they are required to pay less hours. But since customers pay you for hours, it can be difficult to find money to invest in innovation in order to increase automation and increase productivity. From our point of view, since we sell the result for a fixed price, the main interest is to automate everything as much as possible in order to reduce human costs. This allows us to seriously invest in automation . This is beneficial to the client, because systems that are well automated are more stable. The computer makes fewer mistakes than people. This is especially true of large systems and large teams.
I think a large proportion of our success is associated with this. This allows us to be much more flexible in terms of staff. Suppose you sell some billing to a client, and you have some super-developer in your staff that is five times more productive, and therefore you would like to take three times more money for it. In fact, the consulting business model does not work that way. Corporate customers are negatively disposed to a large range in prices, although it is widely known that productivity in the field of software development between middle and high-end specialists may differ tenfold. Therefore, focusing on our subscription model, we can afford to invest in highly efficient personnel .
PG Day: You started talking about corporate-level customers. I think it’s no secret that you have a lot of very large clients like BBC, Airbnb, Cisco and so on. What are the main difficulties of interaction with a client of this level?
Peter: Large companies have higher requirements for structure and consistent results . This was for me one of the challenges in the transformation “Perkony”. When we started, we worked with many startups. Interesting, active guys who could say "Listen, how cool: I ask Perkone a question and get three different opinions from them that have a right to exist." We all understand that in the field of technology there is no one correct solution, as in the problem of mathematics. Usually there are several, and all are quite acceptable: you can do this or that, and everything will work. At first we hired such smart, independent guys and said: “Guys, do what you want. The main thing is that it is reasonable, and the client was satisfied. "
When we started working with enterprise clients, expectations began to change . They want consistency - like, for example, in Starbucks, who say: “Guys, we don’t need the best coffee in the world, but we want to come in, ask for a cappuccino, and may it always be absolutely the same.” Corporations pay a lot of attention to this aspect. They want to ask the same question (which they often do because they have many teams that are facing the same problems) and expect to receive the same answer. If they get a different answer each time, it just drives them crazy. And this requires a completely different approach to team organization . , , , , “” . 25 , , . , . enterprise- .
PG Day: ? , . , , .
: , , — . : , , . . - -, , . , . , . — , . : «, SQL . ». , - , MySQL , .
? , . , , . , , . . : , . , . , , : « ?» , , , . “”, , , : “” , . , , , . .
95% . , , . , , , , , , , , , . Joke, of course. , , , . , .
. : , . , , , 5 , , , . , , . -, , . , “”, : «, . , , , !» , . , , . -, , , . .
PG Day: ? DBA — , “” , - .
: , , . , - , . , , , , . , . , background check : - , , -. . , 10 . - , - . , , , - — .
PG Day: , , , -, PG Day'17.
: -, InnoDB. , storage engine, , , 99% MySQL . Percona Live , . : MySQL, - . , . . - , InnoDB: , , , , . , , InnoDB , , “” - , .
PG Day: Who can be interested in participating in a master class in the first place?
Peter: Most of all this master class is interesting DBA and system administrators who work with MySQL. They need to know how to use the maximum MySQL capabilities. Listeners will have many chances to ask questions, discuss some of their applied problems in the context of the material being studied.
PG Day: Thank you, Peter!
Source: https://habr.com/ru/post/330716/
All Articles