-
-
Notifications
You must be signed in to change notification settings - Fork 1.8k
Error during 0024_auto_20191230_2052 migration #322
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
Oh, this is probably something we can protect against but I need our expert on this: @lynnagara - any ideas? |
I can't reproduce this anymore for some reason. This worked the same way yesterday when I tested this multiple times. Now I always get different error running same commands: sudo rm -rf /var/lib/docker/volumes/sentry-postgres
sudo cp ~/sentry-postgres /var/lib/docker/volumes/ -r
./install.sh gives me Faking migration for contenttypes.0001_initial
Faking migration for contenttypes.0002_remove_content_type_name
Faking migration for auth.0001_initial
Faking migration for auth.0002_alter_permission_name_max_length
Faking migration for auth.0003_alter_user_email_max_length
Faking migration for auth.0004_alter_user_username_opts
Faking migration for auth.0005_alter_user_last_login_null
Faking migration for auth.0006_require_contenttypes_0002
Faking migration for sites.0001_initial
Faking migration for admin.0001_initial
Faking migration for sessions.0001_initial
Traceback (most recent call last):
File "/usr/local/bin/sentry", line 8, in <module>
sys.exit(main())
File "/usr/local/lib/python2.7/site-packages/sentry/runner/__init__.py", line 164, in main
cli(prog_name=get_prog(), obj={}, max_content_width=100)
File "/usr/local/lib/python2.7/site-packages/click/core.py", line 722, in __call__
return self.main(*args, **kwargs)
File "/usr/local/lib/python2.7/site-packages/click/core.py", line 697, in main
rv = self.invoke(ctx)
File "/usr/local/lib/python2.7/site-packages/click/core.py", line 1066, in invoke
return _process_result(sub_ctx.command.invoke(sub_ctx))
File "/usr/local/lib/python2.7/site-packages/click/core.py", line 895, in invoke
return ctx.invoke(self.callback, **ctx.params)
File "/usr/local/lib/python2.7/site-packages/click/core.py", line 535, in invoke
return callback(*args, **kwargs)
File "/usr/local/lib/python2.7/site-packages/click/decorators.py", line 17, in new_func
return f(get_current_context(), *args, **kwargs)
File "/usr/local/lib/python2.7/site-packages/sentry/runner/decorators.py", line 30, in inner
return ctx.invoke(f, *args, **kwargs)
File "/usr/local/lib/python2.7/site-packages/click/core.py", line 535, in invoke
return callback(*args, **kwargs)
File "/usr/local/lib/python2.7/site-packages/click/decorators.py", line 17, in new_func
return f(get_current_context(), *args, **kwargs)
File "/usr/local/lib/python2.7/site-packages/sentry/runner/commands/upgrade.py", line 168, in upgrade
_upgrade(not noinput, traceback, verbosity, not no_repair)
File "/usr/local/lib/python2.7/site-packages/sentry/runner/commands/upgrade.py", line 130, in _upgrade
ignore_ghost_migrations=True,
File "/usr/local/lib/python2.7/site-packages/django/core/management/__init__.py", line 130, in call_command
return command.execute(*args, **defaults)
File "/usr/local/lib/python2.7/site-packages/django/core/management/base.py", line 345, in execute
output = self.handle(*args, **options)
File "/usr/local/lib/python2.7/site-packages/django/core/management/commands/migrate.py", line 86, in handle
executor.loader.check_consistent_history(connection)
File "/usr/local/lib/python2.7/site-packages/django/db/migrations/loader.py", line 292, in check_consistent_history
connection.alias,
django.db.migrations.exceptions.InconsistentMigrationHistory: Migration jira_ac.0001_initial is applied before its dependency sentry.0003_auto_20191022_0122 on database 'default'. every time. Database backup was not changed, sentry 9.12 runs with it without problems. |
@Fogapod, yes, can you file a new issue for this second one. I'm fairly sure it is due to a missing line over at https://github.com/getsentry/sentry/blob/2ee72b121658496633e71d77f34e006de88a220b/src/sentry/conf/server.py#L1739-L1741 and I hope to resolve this one very soon with the help of @wedamija. I'm guessing you had the Jira plugin installed earlier? (or just sentry-plugins?) |
I'm nots sure, I had github plugin installed before but I'm not using it anymore, it might got uninstalled at some point |
All plugins are now baked in so no need to install them anymore and that would explain the issue as at some point you had the Jira plugin installed too (as part of the |
* ref(ts): Convert <SidebarOrgSummary> to typescript * comment * fix: Use django's is_safe_url function (#16337) The implementation of this function from django 1.9.13 covers more invalid URLs than the one from django-sudo does. We should not redirect to `http:123347` style URLs. * fix(ui) Fix gravatar fallbacks (#16336) When updating Gravatar to typescript, I removed some property spreads as they triggered typescript warnings and the propTypes didn't cover any props that needed spreading. The onLoad and onError props were load bearing as they are used to handle gravatar fallbacks. * Removing unused class (#16278) * fix(discover) - Show errors if we encounter them loading discover - Shows a permission error if a user without access to events like a billing account tries to access discover * feat(discover) Switch trace and span id to use columns instead of context Have the trace_id and span_id use the built in columns instead of the context array. * fix(ui): Fix <ContextPickerModal> and refactor to ts (#16090) This removes `LatestContext` dependency and fixes `ContextPickerModal` and refactors it to ts. Co-Authored-By: Matej Minar <[email protected]> * fix(ui): Fix redirect after creating/editing Metric Alert rules (#16342) Also makes sure that the index redirects to combined alert listing view. * ref(perf) Attempt to improve performance of facets endpoint (#16359) * Fetch tag values individually. This will let snuba use the promoted columns more effectively as they perform better. * Only get 10 facets by default. * Use a having clause instead of WHERE to exclude tracing tags as it performs better. * Don't use fields that unpack into arrayJoin() expressions where possible. * Enable sampling at lower project thresholds, as we don't need full accuracy. * fix(saml2): Improve error message on attribute mapping (#16347) * ref: Remove SnubaEvent (#16309) Remove SnubaEvent since we are moving towards a single event model (eventstore/models/Event) everywhere. * fix(tasks) Add event data to user report asynchronously If an event hasn't been propagated when the user report is created, then queue a task to sync the event data once it's stored fully. * feat(django): pytest startup and test collection on Django 1.11 (#16105) * chore: cleanup bitfield (#16031) * fix(migrations): Make Snuba event migration more robust (#16364) Fixes getsentry/self-hosted#322 * feat(teams): Update open membership team permissions (#16244) * fix(migrations): Fix prerequisite for jira_ac migration (#16351) Fixes getsentry/self-hosted#332 * fix(auth): Correct path used for SSO errors (#16315) The route for sentry-organization-auth-settings was still located in the old settings path, this route is used for a redirect in the auth.helper module when there is a problem configuring SSO. This just corrects the redirect so it goes back to settings instead of 404ing. * fix: Remove beta tag on PagerDuty integration (#16344) * fix(tagstore) Don't explode on null values (#16368) The `sentry:user` tag can come back null when it is read from the promoted columns. Handling nulls allows get_tag_value_label() to be used more broadly. * ref: Remove now unused discover query code (#16306) Remove the automatic dataset detection logic from sentry. We have switched over to the snuba based implementation of this. Not having dataset selection in sentry also means we can remove some (but not all) of the supporting logic as well. * feat(ui): Require `org:write` role for Metric Alerts form (#16340) This disables all form fields when editing a Metric Alert without the `org:write` role. * ref(ts) Move BadgeDisplayName to typescript (#16361) * test: fix remaining Django 1.11 tests and make required (#16367) * fix(saml): Wrap individual idp config tabs with forms (#16373) * chore(django 1.11): bump django-crispy-forms to 1.7.2 (#16371) * translation+pluralization * removed duplicate style * removed duplicate style * fix(ui): Fixed alignment in <TogglableAddress> if it cannot be converted to relative (#16385) * ref(discover2) Test out sampling on tag values (#16387) The facets endpoint is still not fast. I want to see what kind of results can be obtained if we aggressively sample tag values. If this yields acceptable performance I want to try scaling the sampling rate based on the data volume so that we can better balance accurracy and performance. * test(acceptance): Try #2 to fix flakey onboarding test for dja… (#16345) Try #2, use a timed retry policy * fix(discover2): Fix Percy snapshots and UX issues on Querybuilder (#16291) * fix(discover2): Test fix on Percy snapshots * fix(discover2): Improve QueryBuilder for error/loading states * fix(discover2): Resize QueryBuilder when browser is resized * feat: Register option for eventstream (#16363) Register option for #16285 * feat(alerts): Remove "Related issues" from incidents (#16379) This removes the UI components that fetch and render related issues on the Incidents details view. * ref: Deprecate get_legacy_message (#16310) Move the implementation into the Pagerduty plugin since this is the only place it is used. * feat(tracing) Add spans for snuba queries (#16388) Add spans to collect more information about snuba queries. Having the referrer and the start of the query body should make tracking down slow queries in snuba easier as we have more context. * feat(workflow): Remove "Create incident" manual workflow and e… (#16375) This removes the UI and endpoint to create a manual incident. * feat: upgrade drf to 3.6.4 (#16392) * feat(alerts): Remove "Suspects" UI and endpoint (#16378) This was used on the Incidents details view, but is no longer used * ref(discover) Remove the raw field (#16390) Outputting the raw data when we also output all the formatted keys wastes space and adds visual noise. * fix(discover2): Use transaction root span only if there are no descendent spans for the ops breakdown (#16377) * feat(emotion): Upgrade emotion to v10 (#15106) * ref(ui): Refactor `cx` from emotion to `classnames` We generally use the `classnames` library to handle merging/dynamic `className`. No reason to use two different libraries. * upgrade emotion, add styled wrapper for typings (and related paths), add eslint rules for emotion, update babel plugin * Refactor usage of emotion `css` being treated as a class name string * change innerRef --> ref * refactor grid-emotion props and update to `reflexbox` * eslint for emotion-setup * refactor injectGlobal --> Global for emotion upgrade * Add <CacheProvider>, this is needed to turn speedy off for percy * refactor <HeaderItem> - fix missing props to styled components, remove innerRef and forwardRef * Change <Tooltip> logic for when to apply a wrapper, emotion causes the previously expected types to be different * random TS fixes for styled components * fix <RadioGroup> animation needing `css` * forwardRef for Switch * define a `theme.space` for rebass/grid * fix DOM attribute warnings with "loading" prop * fix settings breadcrumb - Box in `reflexbox` has different CSS than grid-emotion * change <ProviderRow> to remove `Flex` around CircleIndicator as it was squishing it on small widths * css fixes for percy * fix featureDisabled logic for custom alerts because emotion * refactor emotion9 --> emotion10 (imports) * update snapshots and fix tests * feat(drf): Add CamelSnakeSerializer (#16397) * docs(workflow): Fix link to screenshots in README.md (#16404) Fixes links to screenshots in README.md broken by 983c8c6 * ref(ts): Convert <QueryCount> to typescript (#16356) * ref(ts): Convert <SearchBar> to typescript (#16357) * ref(ts): Convert <SearchBar> to typescript * removed proptypes, placeholder optional * ref(ts): Convert getOnboardingTasks to typescript (#16348) * merge Co-authored-by: Mark Story <[email protected]> Co-authored-by: Chris Fuller <[email protected]> Co-authored-by: William Mak <[email protected]> Co-authored-by: evanh <[email protected]> Co-authored-by: Billy Vong <[email protected]> Co-authored-by: Evan Purkhiser <[email protected]> Co-authored-by: Lyn Nagara <[email protected]> Co-authored-by: josh <[email protected]> Co-authored-by: Burak Yigit Kaya <[email protected]> Co-authored-by: Megan Heskett <[email protected]> Co-authored-by: NisanthanNanthakumar <[email protected]> Co-authored-by: Danny Lee <[email protected]> Co-authored-by: Alberto Leal <[email protected]> Co-authored-by: Aakash Kambuj <[email protected]>
I have fixed first reported error directly deleting a few events referencing not existing groups:
|
I tried upgrading my 9.12 installation to 10 and this error happened during database migration:
Looks like my database is a bit broken, one project got deleted, but there are still references to it from some tables. I had an error in previous migration too and had to delete some rows from tables that pointed to deleted
project_id
. This error is probably related to that too, but I don't understand what exactly causes it and what can I do.How can I fix this?
I have a backup of my database volume and can try different solutions.
The text was updated successfully, but these errors were encountered: