On the morning of May 16, Beijing time, and the evening of May 15, American Eastern time, Bruce Momjian, co-founder of PostgreSQL international community and vice president of EDB, visited the activity site of “PostgreSQL celebrities face to face” of China’s PostgreSQL branch. With the help of simultaneous interpretation of Cary Huang and David Zhang, nearly 100 PostgreSQL practitioners and enthusiasts discussed the future PostgreSQL segments.
No.1 technology sharing
In the lecture session, Bruce Momjian shared with you what is database extension; why to expand; options and methods of vertical expansion; options and methods of non slicing horizontal expansion; significance of using fragmentation; existing slicing options; future fragmentation requirements; parallel data sharding access, etc., which are rich, wonderful and nutritious.
No.2 Q & A interaction
In the subsequent Q & a interactive session, according to the questions raised by everyone on the spot in the chat bar, Bruce Momjian was happy to answer them one by one. Bruce Momjian revealed that PostgreSQL 13 will be released in a few weeks, but there is not a lot of slicing management added. “The lack of demand for fragmented business scenarios is the reason for the slow development of the community, but it will not make people wait too long, because the native distributed function will be released as a heavyweight of PostgreSQL 14.”
Bruce Momjian highly recognizes the contribution of Chinese enterprises to sharding in the international community. For example, highgo has submitted many patches on sharding for the community, including parallel external scanning of FDW, snapshot based on global CSN and Si integration. It is also hoped that PostgreSQL 14 can include more contributions from China.
As there are many on-site questions, Bruce Momjian’s answers are not shown here.
Summary of field problems:
- Which version of PG native cluster can be released?
- Where does metadata exist? If there is a coordinator, how to deal with the problem that the coordinator can not be synchronized due to downtime? Can you consider putting metadata on a separate raft replication group?
- What are the advantages of FDW and citus?
- If the PostgreSQL community wants to implement sharding, is there a major design different from PostgreSQL XL?
- The xid64 community has submitted the patch very early, and the dreamp engine has also submitted it very early. When can the long-standing mvcc problem be solved?
- At present, in some scenarios, the performance of global sequences will be greatly reduced in the distributed environment. What optimization methods are there for PG’s follow-up plans?
- Will GTM become a bottleneck? How to solve this problem? If the partition number is modified, is it necessary to re import the data instead of importing the data? Because the fragmentation maintenance is more complex than the non fragmentation maintenance, will there be a matching better maintenance tool for maintaining the fragmentation?
- For distributed keys without uniqueness constraints, hash algorithm can not solve the problem of data skew. What distribution algorithms do PG have at present? And how to avoid the problem of sliced data skew?
- Now the new-type newsql generally adopts the raft like replication mode. Is there any plan to use it after PG?
- Is there any plan to support listing in the future of PG? What is the progress of this work?
- Is there any component in PG that supports data compression? Now some historical archives take up too much space. Is there any plan to support this community?
No.3 dialogue with Lianmai
At the end of the dialogue with Lianmai, Bruce Momjian answered “why he always loves technology and PostgreSQL and some thoughts.” The following is the general content of the dialogue:
ask: in China, many people of your age are thinking about retirement. What makes you so keen on technology and PostgreSQL database?
answerOne is interest. When I contacted the database, I was very young at that time. At that time, I heard about encryption, hash and other terms, but I didn’t know how to work, so I decided to study it in my spare time, including some basic research. Interest has always driven me forward. The two are experiences. When I was young, I served as a teacher for five years. The teacher’s experience made me form the habit of continuous learning. The three are habits. Whether I work at EDB, the international community or at home, I will never lose interest and develop into a habit. That’s why I keep loving and passionate.
Finally, I would like to express my sincere thanks to Bruce Momjian for participating in the “face-to-face PostgreSQL celebrities” activity invited by China’s PostgreSQL branch. Secondly, I would like to thank volunteers Cary Huang and David Zhang for their accurate Chinese and English real-time translation. Finally, I would like to thank PostgreSQL lovers and followers from all walks of life for their participation and support from PostgreSQL practitioners.
Bruce Momjian blog address:
Active playback video address:
Ppt acquisition method:
Concerned about the official WeChat official account of China PostgreSQL branch and reply to “big coffee first issue”.
For more wonderful content, please pay attention to the following platforms and websites:
Official account of China Postgre SQL chapter (technical articles and technical activities):
PostgreSQL branch of Open Source Software Alliance
China postgre SQL branch technical Q & a community:
Official website of China postgre SQL branch: