Is it necessary to configure JPA maxHistoryRecordNum in application.yml?

[LL-LIN]

I believe that only the time window and the exceptions of the database system itself are needed to control the amount of data.
For example, the following code

            try {
                int rows = historyDao.deleteHistoriesByTimeBefore(expireTime);
                log.info("[jpa-metrics-store]-delete {} rows.", rows);
                long total = historyDao.count();
                if (total > jpaProperties.maxHistoryRecordNum()) {
                    rows = historyDao.deleteOlderHistoriesRecord(jpaProperties.maxHistoryRecordNum() / 2);
                    log.warn("[jpa-metrics-store]-force delete {} rows due too many. Please use time series db instead of jpa for better performance.", rows);
                }
            } catch (Exception e) {
                log.error("expiredDataCleaner database error: {}.", e.getMessage());
                log.error("try to truncate table hzb_history. Please use time series db instead of jpa for better performance.");
                historyDao.truncateTable();
            }

[LL-LIN]

Due to the lack of configuration for maxHistoryRecordNum (using the default value), the historical data has not been displayed correctly, and the monitoring server downtime has not been updated in a timely manner, which has confused me for a long time.

The version I use in my production environment is 1.6.1, and I see that the logic in the master branch is also consistent.