I have a question about the postgresql crash. By default, kubernetes, help installation, and postgresql are used The conf configuration is also default. The number of users is about 50. 500 errors often occur. Using a single node, the running memory is 32GB. An error occurs: PANIC: could not fdatasync file “000000000000100000010000004B”: Interrupted system call, as shown in the following figure. What is the reason?

This problem has been improved on Kubernetes because

  1. I use network storage NFS or other
  2. redis persistent storage is switched to rdb (to prevent continuous read and write)
  3. Limit the postgresql parameter disk reading (pay attention to prevent data loss)