* run annotation data migration in batches
* how could i miss it
* run in the background, starting from newest annotations
* update tests
* optionally pass batch size via env
* fix: collate rule_group column as utf8mb4_bin
This ensures this column is sorted identically by the database and
golang, which should eliminate issues related to a mismatch in sort
order.
* chore: make test stricter
* fix: also add binary collation for postgres
This adds a `message` column to the `alert_rule_version` table. This follows the
pattern established for dashboards as closely as possible. A new type is
introduced internally for passing the new `message` field around in a type-safe
manner, but doing the same for the API types becomes very messy. In that case, a
new field is added with omitempty.
Note this PR is only:
- The column addition
- The "read" path; API for listing versions
Subsequent PRs will add code to actually set the message when updating rules.
The query which fetches alert rules in a paginated manner ordered by `rule_group` can result in strange and inconsistent ordering when the database uses a case-insensitive collation for the `rule_group` column. This can lead to scenarios where rules from different groups are interleaved in the results, making pagination unreliable and the returned number of rule_groups incorrect.
Related to #88990
* Convert unique keys in file and file_meta tables into primary key.
* Fix panic.
* Fix comment.
* Always add migration to drop auto-generated PK.
* Drop and create PK for mysql in single statement.
* Drop my_row_id column too. (Please drop primary key column to be able to drop generated invisible primary key.)
* Cloud migrations: store snapshots in the database
* update github.com/grafana/grafana-cloud-migration-snapshot to v1.9.0
* make update-workspace
* use new field name in test
* return error after call to fmt.Errorf
* create methods for readability / fix session deletiong not deleting snapshots
* remove debugging changes
* update sample.ini
* update tests to include OrgID in ListSnapshotsQuery
* lint
* lint
* Update pkg/services/cloudmigration/cloudmigrationimpl/snapshot_mgmt.go
Co-authored-by: Matheus Macabu <macabu@users.noreply.github.com>
* remove TODO
* Update pkg/services/cloudmigration/cloudmigrationimpl/snapshot_mgmt.go
Co-authored-by: Matheus Macabu <macabu@users.noreply.github.com>
* remove one of the debug logs
---------
Co-authored-by: Matheus Macabu <macabu@users.noreply.github.com>
* automatically rename integration tests to follow the common convention
* name tests differently
* alter column type to bigint
* update another column to bigint
* add another alter
* fix subquery for mysql
* fix: increase login_attempt.ip_address column length for IPv6 support
- Expand ip_address column from VARCHAR(30) to VARCHAR(50) to accommodate IPv6 addresses
- Add database migration with support for PostgreSQL, MySQL, and SQLite
- Add comprehensive integration tests for various IPv6 address formats
- Resolves 500 errors when login fails over IPv6, now returns proper 401 errors
Fixes#106362
* test: add missing test skip to TestIntegrationIPv6AddressSupport
Skip integration test when running with -short flag to separate unit and integration tests
* Update pkg/services/sqlstore/migrations/login_attempt_mig.go
Co-authored-by: Victor Cinaglia <victor@grafana.com>
* fix missing bracket
* fix: resolve PostgreSQL timestamp overflow in IPv6 test
- Use controlled time mock instead of time.Now() to avoid timestamp conversion issues
- Follow existing test patterns with xormStore and mock time functions
- Add proper Since parameter to GetIPLoginAttemptCount query
- Fixes PostgreSQL error: 'pq: value "-62135596800" is out of range for type integer'
* fix: resolve PostgreSQL UTF-8 encoding error in IPv6 test
Replace string(rune(i)) with fmt.Sprintf to avoid null bytes (0x00)
when i=0, which caused 'invalid byte sequence for encoding UTF8' error
---------
Co-authored-by: Victor Cinaglia <victor@grafana.com>
* replace xorm.io/xorm imports
* replace xorm from other go.mod files
* clean up workspace
* nolint does not make sense anymore as it is not a module
* try if nolint directive helps
* use nolint:all for xorm
* add more nolints
* try to skip xorm in linter config
* exclude xorm differently
* retrigger ci
* add columns external_id and is_provisioned to the team table
* generate openapi specs
* rename column to external_uid
* generate open api specs
* increase limit for external_uid to 256
* CI: Allow Bench conversion to fail
We shouldn't mark PRs and commits as X if they fail to convert logs with Bench.
* Fix: Prints should always include new lines
* fix: remove unused import
* Validate authID when user is provisioned
* Add new `user_unique_id` to `user_auth` table
* Validate provisioned user with saml assertion
* Rework `ExternalUID`
* Validate for ExternalUID only
* Enhance verbosity
* Move ExternalUID to saml config
* Rename db variable for externalUID
* Add verbosity to debug ExternalUID
* Assign new error for ExternalUID mismatch
* Add `GetByLoginFn`
* Add new configuration to saml tests
* add validation for empty externalUID
* Fix UNION syntax in resourcepermissions package.
* Fix migrations in usermig package to work with Spanner.
* Fix health query.
* Use more connections for integration tests.
* Add test-go-integration-spanner target to run integration tests against Spanner emulator.
* Add test for enterprise.
* Don't delete sequence number for migration_log.id column.
* Only bump max open connections to 20 for Spanner.
Lower integration test timeout.
* Add migration to remove datasources:drilldown
* remove role assignments as well
* rename to datasources:drilldown
* remove the role assignments
* made it simple instead
* Update pkg/services/sqlstore/migrations/accesscontrol/datasource_drilldown_removal.go
Co-authored-by: Gabriel MABILLE <gamab@users.noreply.github.com>
* Update pkg/services/sqlstore/migrations/accesscontrol/datasource_drilldown_removal.go
Co-authored-by: Gabriel MABILLE <gamab@users.noreply.github.com>
---------
Co-authored-by: Gabriel MABILLE <gamab@users.noreply.github.com>
What is this feature?
This PR introduces a new alert rule configuration option, keep_firing_for (Prometheus documentation).
keep_firing_for prevents alerts from resolving immediately after the alert condition returns to normal. Instead, they transition into a "Recovering" state and are not considered resolved by the Alertmanager. Once the recovery period ends (or after the next evaluation if it is bigger than keep_firing_for), the alert transitions to "Normal" if it doesn't start alerting again:
Before
+----------+ +----------+
| Alerting |---->| Normal |
+----------+ +----------+
-----
After
+----------+ +------------+ +----------+
| Alerting |----->| Recovering |---->| Normal |
+----------+ +------------+ +----------+
Why do we need this feature?
This feature prevents flapping alerts by adding a recovery period. This helps avoid false resolutions caused by brief alert
