Rename the field 'osv_memory' on 'ir.model' into 'transient', and make it a
regular field instead of a computed one. When instantiating a custom model,
use its value to make transient models.
This commit alone will make your server crash on an existing database. A small
database migration is needed: simply add a boolean column `transient` (with
default value false) in table `ir_model`.
- Preserved explicit 3rd-party copyright notices
- Explicit boilerplate should not be necessary - copyright law applies
automatically in all countries thanks to Berne Convention + WTO rules,
and a reference to the applicable license is clear enough.
The patch speeds up the cache invalidation in the registry, and make most
models use the generic method `clear_caches` instead of specific `clear_cache`
methods.
The lazy property `pure_function_fields` was not invalidated upon every setup
of models, and hence could contain old instances of fields. As every model
setup re-creates instances of fields, the property has to be recomputed.
A cross-registry cache was introduced by e2ea691ce.
The initial idea was praiseworthy but sub-optimal for servers with a
lot of registries.
When there is lot of registry loaded, the cache size was huge and
clearing some entries took a lot of CPU time, increasing the chances
of timeout.
Also, the cache was not cleaned when a registry is removed from
registry LRU (this operation would also consume time).
The registry size is now assumed to be around 10Mb, and ormcache size is
proportional to the maximum number of registries. Statistics about ormcache
usage is shown to the log when receiving signal SIGUSR1.
The ormcache is now shared among registries. The cached methods use keys like
(DBNAME, MODELNAME, METHOD, args...). This allows registries with high load to
use more cache than other registries.
The model setup sometimes misses entries in _inherit_fields and _all_columns.
This is because those dictionaries are computed from parent models which are
not guaranteed to be completely set up: sometimes a parent field is only
partially set up, and columns are missing (they are generated from fields after
their setup).
To avoid this bug, the setup has been split in three phases:
(1) determine all inherited and custom fields on models;
(2) setup fields, except for recomputation triggers, and generate columns;
(3) add recomputation triggers and complete the setup of the model.
Making these three phases explicit brings good invariants:
- when setting up a field, all models know all their fields;
- when adding recomputation triggers, you know that fields have been set up.
Unify and refactor exception handling in framework and addons.
The generic `except_osv` is now deprecated, and replaced by more specialized exception subtypes:
- `UserError` (renamed from Warning, as it conflicts with the built-in `Warning`) raised when a non-technical error occurs during a business operation. It could be a missing information in the data provided by the user, or a misconfiguration.
- `AccessError`: raised when any operation is denied because the user conducting it does not have the required access rights.
- `AccessDenied`: raised when an operation that requires authenticated access is attempted via an unauthenticated request.
- `MissingError`: raised when an operation is attempted on a record that does not exist.
- `ValidationError`: raised when an operation violates a SQL or Python constraint.
- All other exceptions are internal errors due to a system problem or bug, and raised untouched to the client-side, which should display a traceback.
All exceptions take a single message argument.
The `test_exceptions` module has been updated to showcase both new and old (deprecated) exceptions.
A great many old `except_osv` had a useless title with "Error!" or "Warning", those have been removed, as this is handled by the client-side widget that displays the messages.
This commit introduces a more consistent policy for logging errors and warnings:
- All messages that do not require administrator attention should be logged at INFO level or lower. This includes all errors that are notified to the user in a friendly manner, even for access right problems or validation errors during business operations.
- All messages that indicate a likely misconfiguration or malicious use by the users should be logged at WARNING level, as they typically require administrator attention.
- All other unhandled internal errors cannot typically be handled by the user and should be logged at ERROR or higher level, as they require immediate administrator attention.
* document and warn that checks and fast_suite in tests sub-packages are
deprecated and have no effect
* avoid iterating all currently loaded modules when looking for test
modules in a tests sub-package
* replace use of __import__ by importlib
Fixes#3152
When loading the registry without any module installation/upgrade, models are
set up once instead of twice. In other cases, models are always set up before
installations/upgrades.
Loading views for custom models from module data files was not possible because
custom models and fields were introduced into the registry after all modules
were loaded. As a consequence, the view architecture did not pass the checks.
This patch takes a different approach: custom models and fields are loaded
early on in the registry, so that views can be validated. The trick is to take
special care of relational custom fields: we skip them if their comodel does
not appear in the registry. This allows to install and upgrade modules that
create/modify custom models, fields and views for them.
This simple optimization in load_marked_modules() avoids unnecessary calls to
load_module_graph(). This provides a small speedup, and avoids confusing log:
some module updates were making it look like the registry was loaded 5 times
instead of once.
Changing the decimal precision of float fields is a rare
operation, while cache clearing occurs fairly frequently.
Signaling a full registry change when the decimal precision
is changed (instead of a mere cache change) is therefore
a better trade-off, and more semantically correct as well.
This way we avoid the decimal precision refresh for each
invalidation.
Registry invalidation implies cache invalidation.
There was an issue in _setup_fields(): the method invokes _inherits_reload(),
which recomputes inherited fields, and invokes itself recursively on children
models. This may be problematic if the children models have already been set
up.
This optimization avoids recursive calls of method _inherits_reload(). In
_setup_fields(), first all parent models are set up, then their fields are
inspected to determine inherited fields, and their setup is done. This scheme
guarantees that inherited fields are computed once per model.
When working with a large number of databases, the memory allocated to
registries wasn't limited, resulting to waste memory (especially in the
longpolling worker, which is not recycled).
The size of the LRU is depending on the soft limit configured for
workers.
Indeed using fromkeys with a list / dict as argument leads to the creation
of shared list / dict. This could create some ugly side effects when
used in loops. This commit fixes or cleans this kind of statement to avoid
unwanted side effects.
When a decimal_precision record is created/modified, the float fields of the
models in the registry must be reset. This was done on old-API columns only.
It is now handled by the new-API fields.
If the server was started without -i or -u and
happened to initialize a fresh database,
auto-installed modules that depend
on `base` only would stay in status "to install"
without actually being installed (until the next
installation round was triggered).
This was of little consequence in 7.0, but causes
a crash in 8.0.
Fixes#953
Thomas Groutars