You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Upon every startup, typeorm does the same alter. Over and over again. Same alter upon each startup:
ALTER TABLE `push_log` CHANGE `created` `created` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP
Expected Behavior
No alters upon startup when its not necessary
Actual Behavior
Upon every startup, typeorm does the same alter. Over and over again. Same alter upon each startup.
removing @Index(["created"], { unique: false }) helped.
I could reproduce it both on heroku and my laptop ubuntu linux.
I don't know if it was introduced in typeorm only recently, or it was always there and I only noticed because production table grew big
and alter started taking minutes and I actually noticed it.
…ies during synchronization (#7517)
* fix#3991
* fix#3991
* improved test for #2737;
fixed remaining issues in #2737;
* fix#6412
* added test for #4281 and #4658
* fixed `NOW()` function usage in MySQL/MariaDB;
fixed precision in datetime functions for MySQL/MariaDB;
improved #3991 test for MySQL/MariaDB;
* added test for #2333
* added test for #7381
* added "remove" operation check in #4658 test
* fixed precision in datetime functions for PostgreSQl;
improved #3991 test for PostgreSQl;
* improved #3991 test for CockroachDB;
* improved #3991 test for Sqlite;
* improved #3991 test for Oracle and SQLServer;
Issue Description
Upon every startup, typeorm does the same alter. Over and over again. Same alter upon each startup:
Expected Behavior
No alters upon startup when its not necessary
Actual Behavior
Upon every startup, typeorm does the same alter. Over and over again. Same alter upon each startup.
removing
@Index(["created"], { unique: false })
helped.I could reproduce it both on heroku and my laptop ubuntu linux.
I don't know if it was introduced in typeorm only recently, or it was always there and I only noticed because production table grew big
and alter started taking minutes and I actually noticed it.
Steps to Reproduce
In previous to this fix commit It was reproduceable: BlueWallet/GroundControl@7b3b6f3
table that we have:
entity definition:
My Environment
Relevant Database Driver(s)
aurora-data-api
aurora-data-api-pg
better-sqlite3
cockroachdb
cordova
expo
mongodb
mysql
nativescript
oracle
postgres
react-native
sap
sqlite
sqlite-abstract
sqljs
sqlserver
The text was updated successfully, but these errors were encountered: