Запретить логике приложения читать (не обновлять) данные во время пакетных запусков ⇐ JAVA
Запретить логике приложения читать (не обновлять) данные во время пакетных запусков
I have the following situation: my java spring boot application shall perform calculations based on a large amount of some master data. It must be ensured that the master data correspond to a certain state as a unit. It should not happen that at one point in time, one calculation works with master data A, and simultaneously another calculation processes an older state of A.
So, you can think of the batch as big transaction but of course I cannot technically implement in such a way because data set is very huge and DB connection would be kept open far too long.
Other options would be to lock the all business data functionally until the batch is completed. This would cause the users to be unable to perform their work. Another idea is to keep multiple versions of the data and trying to ensure that for each user session all of the required master data are some state of version. (maybe using hibernate envers)
What solutions did I miss? Have you ever encoutered such a problem?
(So far, I didnt pick any batch framework. My options would be spring batch or quartz)
No implemented anything so far.
Источник: https://stackoverflow.com/questions/781 ... batch-runs
I have the following situation: my java spring boot application shall perform calculations based on a large amount of some master data. It must be ensured that the master data correspond to a certain state as a unit. It should not happen that at one point in time, one calculation works with master data A, and simultaneously another calculation processes an older state of A.
So, you can think of the batch as big transaction but of course I cannot technically implement in such a way because data set is very huge and DB connection would be kept open far too long.
Other options would be to lock the all business data functionally until the batch is completed. This would cause the users to be unable to perform their work. Another idea is to keep multiple versions of the data and trying to ensure that for each user session all of the required master data are some state of version. (maybe using hibernate envers)
What solutions did I miss? Have you ever encoutered such a problem?
(So far, I didnt pick any batch framework. My options would be spring batch or quartz)
No implemented anything so far.
Источник: https://stackoverflow.com/questions/781 ... batch-runs
-
- Похожие темы
- Ответы
- Просмотры
- Последнее сообщение
-
-
Как получить данные от пользователя для специальных ручных запусков в Snowpark
Anonymous » » в форуме Python - 0 Ответы
- 12 Просмотры
-
Последнее сообщение Anonymous
-
-
-
Скрипт Python Почему я не могу зарегистрировать его после нескольких запусков?
Anonymous » » в форуме Python - 0 Ответы
- 11 Просмотры
-
Последнее сообщение Anonymous
-
-
-
Найти ключевые слова в ячейках Jupyter и запустить (проблема одновременных запусков)
Anonymous » » в форуме Python - 0 Ответы
- 16 Просмотры
-
Последнее сообщение Anonymous
-