aiidateam / aiida_core / 6219 / 8

74%

develop: 74%

push

travis-ci

<a href="https://github.com/aiidateam/aiida_core/commit/<a class=hub.com/aiidateam/aiida_core/commit/6da891fee53ca6aaaf95930d038d9a29d339f7df">6da891fee<a href="https://github.com/aiidateam/aiida_core/commit/6da891fee53ca6aaaf95930d038d9a29d339f7df">&quot;&gt;Improve the efficiency of `DbLog` migration for Django (#2949)

The old implementation was using Django&amp;#39;s ORM to get the count and rows
of certain rows in the `DbLog` table. However, this was generating
extremely inefficient queries that made the migration run for days on
big databases. A similar problem was encountered with the original
SqlAlchemy implementation. That was solved in an earlier commit
`</a><a class="double-link" href="https://github.com/aiidateam/aiida_core/commit/<a class="double-link" href="https://github.com/aiidateam/aiida_core/commit/80a0aa9117f399a5385a30ba0ce3e2afc7016a4e">80a0aa911</a>">80a0aa911</a><a href="https://github.com/aiidateam/aiida_core/commit/6da891fee53ca6aaaf95930d038d9a29d339f7df">`, which created a more
efficient query. We now replace the Django ORM solution with the same
custom SQL query as used for SqlAlchemy. This should significantly
improve the efficiency of this migration.

22586 of 32591 relevant lines covered (69.3%)

0.69 hits per line