smtplib.SMTP API calls such as .sendmail() or .login() make sure on
their first line that this command is sent, by calling
.ehlo_or_helo_if_needed().
However, test_smtp_connection() does not use SMTP.sendmail() to simulate
sending email, or the email would really be sent.
Instead it uses the low-level API SMTP.mail() which does not make sure
that HELO was sent.
The connection test could be fixed by ensuring that this command was
sent to the server in the test method. However, this could lead to
inconsistent cases where the connection test passes whereas another
usage in the code fails because the command was not sent.
For example, someone could use the low-level API for some reason.
EHLO could have already been sent because if authentication is enabled
smtplib.SMTP.login() calls ehlo_or_helo_if_needed(). Therefore it is
correct to call it ourselves at the end of our connect() method.
STARTTLS sends a first EHLO, negociates the encryption, and leaves the
state without the second EHLO, which should still be sent to the server,
as stated by RFC 3207, in section "4.2 Result of the STARTTLS Command".
https://www.ietf.org/rfc/rfc3207.txtclosesodoo/odoo#34550
Signed-off-by: Julien Legros (jle) <jle@odoo.com>
It's currently possible for binary downloads to be missing an
extension entirely. Generate an extension from the mimetype and use
that. Also ensure the existing extension (if any) matches the
mimetype.
Warning: this may require adding new mimetype/extensions pairs to the
local mimetype database similar to
f413d155df.
Also fix missing or incorrect filenames on records without a filename
field: in some Odoo versions the client would send a filename of
"null", resulting in a download of e.g. "null.pdf" instead of the
attachment or inferred name.
Task 2025716
closesodoo/odoo#34299
Signed-off-by: Xavier Morel (xmo) <xmo@odoo.com>
A special folder named `0.0.0` can contain scripts that are run on
upgrade of any version.
They are useful to make some sanity checks or other verifications to
ensure database consistency.
The first version of this patch used the more eye-catching `any` for
the migration folder, but it was problematic for upgrading from an older
version that doesn't contain this patch.
Using a version "number" containing two dots is required to avoid it
being prefixed with the server version (see `convert_version` method) and
resulting in a version like `10.0.any`.
Such version would have been executed, even without this patch, when
upgrading from an older major server version (9.0.1.0 < 10.0.any).
closesodoo/odoo#34268
Signed-off-by: Christophe Simonis <chs@odoo.com>
DROP CONSTRAINT (even with IF EXISTS is specified) acquires an ACCESS
EXCLUSIVE lock on the table, preventing e.g. inserts in an other
transaction, so ir_logging would systematically deadlock if configured
to the same database and a warning would be triggered during install
or update (if that ran ir.logging's init).
1. hand-roll the "IF EXISTS" bit, to avoid taking an ACCESS EXCLUSIVE
lock on the table if the problematic constraint does not exist and
thus doesn't need to be dropped (which by now should be the vast
majority of cases).
Replacing DROP CONSTRAINT with DISABLE TRIGGER does not fix the
issue as *that* acquires SHARE ROW EXCLUSIVE. While that's less
constraitning than ACCESS EXCLUSIVE, it still conflicts with an
insert's ROW_EXCLUSIVE.
2. add a timeout to the logging INSERT anyway, the deadlock is still
an issue if we're updating a database which does have the
problematic constraint, and we want to preclude the possible
eventual introduction of new deadlocks in the future.
closesodoo/odoo#34243
Signed-off-by: Xavier Morel (xmo) <xmo@odoo.com>
On `WebRequest` `__exit__`, when an exception occured,
(in `self.registry.signal_changes` or `self.registry.reset_changes`)
cursor were left unclosed as `self._cr.close` was not called
in such cases.
Having exceptions in the above mentioned method do not happen
often, but when it does it left unclosed and unusable cursors
in the connection pool, and in the extreme case explained below,
it left the connection pool with only unclosed and unusable cursors.
The entire server was then unusable as it no longer had working cursors.
Case:
- Start a multi-thread server with db_maxconn set to 5
- Ensure you do not send any request to the server,
not even with a left open tab on `http://localhost:8069` in your browser
- Send 6 parallel HTTP requests to `/web/login`
thanks to an external thread python script
(See below, at the end of this long commit message)
According to your registry state (if you have a lot of modules installed or not),
and the native Python Garbage Collecting state,
you might end with
- either warnings telling some unclosed cursor were garbage collected,
and therefore closed (by a kind of luck thanks to the Python garbage collecting),
- either, a server completely blocked not accepting any other request
(you can try for instance `curl http://localhost:8069`
and you end up with a `500 Internal Server Error`
This observed issue looks to appear only in 11.0. Not 10.0 or 12.0.
This is because only 11.0 clear the cache during registry loading:
`https://github.com/odoo/odoo/blob/f1706c848d41c47646dabca771996e9b9f788241/odoo/modules/loading.py#L236`
This cache clearing doesn't happen in 10.0 nor 12.0
(in 12.0, thanks to e181f592f3)
When sending the 6 parallel requests,
it uses instantly all the 5 available cursors of the connection pool to handle these requests,
and when each request exits, in `__exit__`, it calls `self.registry.signal_changes()`
which tries to open a new cursor because of
- `self.cache_invalidated` which is True, for all the 6 requests, thanks to the call to `clear_caches`
explained above during the registry loading and the fact all requests have been treated in parallel,
- `with closing(self.cursor()) as cr:`, `self.cursor()` attempting to use a new cursor
(the `closing(...)` does not have any incidence on this issue, despite it could look like guilty)
The attempt to use a new cursor fails, as there is no more available (`db_maxconn` is reached),
raising a `PoolError('The Connection Pool Is Full')` exception.
In the request `__exit__` method, because of this exception raised when calling `signal_changes`,
`self._cr.close` is never reached, and the parallel request therefore left only unclosed
cursors in the connection pool,
therefore leaving the server in a state where it only has unusable cursors
and therefore can't do anything more.
This might look like really bad luck to land in such a state,
but we observed multiple actual case on Odoo.sh,
the one referenced in this commit (opw-2008340) was because of an Outlook client
which launched 18 parallel requests to fetch the email images,
and the server wasn't spawned, therefore neither was the registry.
The server registry was therefore just loaded when it received the 18 parallel requests,
and it therefore triggered this extreme use case.
The server was left unusable for several minutes, until a forced restart.
For reference, here is the script that has been used to trigger the 6 parallel requests:
```
import requests
import threading
threads = []
for i in range(6):
threads.append(threading.Thread(target=lambda: requests.get('http://localhost:8069/web/login')))
for thread in threads:
thread.start()
```
opw-2008340
closesodoo/odoo#34071
Signed-off-by: Denis Ledoux <beledouxdenis@users.noreply.github.com>
Installing test_new_api without demo will fail since demo_user
does not exists. This is a problem if we want to test migration
with all modules installed. This simple fix solve the problem.
closesodoo/odoo#34065
Signed-off-by: Christophe Simonis <chs@odoo.com>
Display extra information on error messages:
- Access error based on ACLs: user, fields (if applies)
- Access error based on record rules: user, ids
- Missing error: model, operation, user, ids
closesodoo/odoo#33942
Signed-off-by: Nicolas Martinelli (nim) <nim@odoo.com>
Co-authored-by: jev-odoo <jev@odoo.com>
Backport of fa39019ee1.
If crop_image is called without a default format, it would try the default,
'PNG', to save the result and crash at save when processing non-PNG images.
For example, if a jpeg image was in CMYK colors.
Here it will still crash non-gracefully when failing to save, but the default
should almost always work; if it doesn't it means we failed to detect the format
of the stream, so it is unlikely that we could save the image anyway.
opw 2009029
closesodoo/odoo#34088
Signed-off-by: Nans Lefebvre (len) <len@odoo.com>
Display extra information on error messages:
- Access error based on ACLs: user, fields (if applies)
- Access error based on record rules: user, ids
- Missing error: model, operation, user, ids
closesodoo/odoo#33945
Signed-off-by: Nicolas Martinelli (nim) <nim@odoo.com>
Co-authored-by: jev-odoo <jev@odoo.com>
Before this task, when creating the related field from the studio, it is by
default stored. When we change the field name in 'ir.model.fields' using ORM's
write method, it also renames the column in the table associated with the
field's model.
This is the expected behavior. But when someone tries to change the 'name' of a
non stored field, the column is not available in the associated model, and so
trying to rename the column results into traceback. This issue was probably
there for a while but came into light with the task 1985710. More info can
be found on the chatter of the task.
This commit fixes the issue by adding an extra check before renaming the table
column. The query will be executed only when the field is stored and thus
exists in the table.
closesodoo/odoo#33944
Signed-off-by: Raphael Collet (rco) <rco@openerp.com>
The content of these fields can't be updated, there's no reason to
even show them. Just remove them from the default / auto-generated
view for client actions.
Task 1772242
closesodoo/odoo#29179
Signed-off-by: Xavier Morel (xmo) <xmo@odoo.com>
Since test_pylint was moved into the test_lint module in commit
b27077bb, this could break Odoo instances where a module has test_pylint in its dependency.
With this commit, the test_pylint module is coming back as an empty
module.
closesodoo/odoo#33871
Signed-off-by: Olivier Dony (odo) <odo@openerp.com>
Before this commit:
- When you create a compute field, from the model view, dependence field
was not required when creating a compute field.
- When removing the compute code, the dependence field stays marked as
required.
fine-tuning : f01c195288
opw-2003784
closesodoo/odoo#33760
Signed-off-by: Jorge Pinna Puissant (jpp) <jpp@odoo.com>
Because Odoo still supports Android Kitkat and Internet Explorer, ES5 is
the highest ECMAScript allowed version.
This commit adds a new test that parses each javascript file found in
addons. The test fails if unsupported js code is found. As ES6 is
allowed during tests, the files under a path that contains 'static/test'
are just skipped.
This test uses the es-check tool
(https://www.npmjs.com/package/es-check).
If the es-check tool cannot be found, the test is simply skipped.
Also the test_pylint module is renamed to a more generic name test_lint,
that way, every linter test can find a shelter here.
closesodoo/odoo#33724
Signed-off-by: Christophe Monniez (moc) <moc@odoo.com>
Paul Morelle