Worker is now the base class of all workers. It has a message
queue and can be run in a thread of its own, or in the calling
thread. Worker can not be used as such, but a concrete worker
class must be derived from it. Currently there is only one
concrete class RoutingWorker.
There is some overlapping in functionality between Worker and
RoutingWorker, as there is e.g. a need for broadcasting a
message to all routing workers, but not to other workers.
Currently other workers can not be created as the array for
holding the pointers to the workers is exactly as large as
there will be RoutingWorkers. That will be changed so that
the maximum number of threads is hardwired to some ridiculous
value such as 128. That's the first step in the path towards
a situation where the number of worker threads can be changed
at runtime.
A new class mxs::Worker will be introduced and mxs::RoutingWorker
will be inherited from that. mxs::Worker will basically only be a
thread with a message-loop.
Once available, all current non-worker threads (but the one
implicitly created by microhttpd) can be creating by inheriting
from that; in practice that means the housekeeping thread, all
monitor threads and possibly the logging thread.
The benefit of this arrangement is that there then will be a general
mechanism for cross thread communication without having to use any
shared data structures.
The output generated by a failed call to a module command was previously
overwritten with the error messages stored in the module command
subsystem.
In the case of a failure, the proper procedure is to check if the output
generated by the module command conforms to the JSON API error
specification and if it does, combine it with any system generated error
messages. If the output does not conform, it is stored in the "meta" field
of the returned object. This allows all of the generated output to be
saved.
All internal code is now inside an anonymous namespace to prevent their
use outside of the compilation unit.
Also fixed the wrong return type of ResourceWatcher::etag.
Executing the commands inside a worker thread allows further improvements
to job queuing but mainly it fixes the problem of loading users when
listeners are allocated at runtime.
When a runtime listener was being created, it was allocated in the admin
thread whereas the listeners created at startup were allocated in the
"main" thread. This caused a minor difference in how administrative
functions were handled by the REST API and MaxAdmin. The only real problem
is that listener allocation depends on being done inside a worker thread
as it lazily initializes some resources.
The internal header directory conflicted with in-source builds causing a
build failure. This is fixed by renaming the internal header directory to
something other than maxscale.
The renaming pointed out a few problems in a couple of source files that
appeared to include internal headers when the headers were in fact public
headers.
Fixed maxctrl in-source builds by making the copying of the sources
optional.
The JSON API specification states that all resources must support direct
modification of resource relationships by providing only the definition
for a particular relationship type to a /:type/:id/relationships/:type
endpoint.
The relevant part of the JSON API specification:
http://jsonapi.org/format/#crud-updating-to-many-relationships
Since the module command interface was expanded to include a JSON output
parameter, there is no longer a need for an output DCB. As the JSON can be
printed by both maxadmin and the REST API, this allows the removal of
explicit output formatting in module commands.
Cleaned up the documentation and reformatted the parameter list of
listener creation to be made out of links.
Added some comments to the resource declarations in resource.cc to prevent
the order of them from being altered.
The resource collection modification times weren't updated when an
individual resource was updated. By tracking back the path of a resource
and updating the modification times, all nodes in the path will get the
correct modification time.
The checks whether a request body is present are now done at a higher
level. This removes the need to do the checks at the resource handler
callback, removing duplicated code.
The checks are done by adding constraints to resources that must be
fulfilled by each request.
Added debug assertions to make sure that the core logic of the REST API
resource system works.
The listeners are now a proper sub-resource of the service resource. This
means that it acts like a normal resource and can be queried both as a
collection of resources and as an individual resource.
All resoures now use the `state` member to describe their internal
state. This includes servers, services and monitors. This means that the
`status` keyword can be reserved for something else and it can be removed
until it is needed again.
Changed the module maturity field to `maturity` to better describe its
purpose.
The module commands can now produce JSON formatted output which is passed
to the caller. The output should conform to the JSON API as closely as
possible.
Currently, the REST API wraps all JSON produced by module commands inside
a meta-object of the following type:
{
"meta": <output of module command>
}
This allows the output to be JSON API conformant without modifying the
modules and allows incremental updates to code.
The PATCH method should now be used instead the PUT method to update
resources. As PUT request bodies should represent complete resources, the
use of PUT to update resources is no longer supported.
Altered tests to use PATCH instead of PUT for updating resources.
The module command self links now point to an endpoint that executes the
module command. Depending on the type of the module command, either a GET
or a POST request must be made.
The /maxscale/ resource now supports PUT requests which modify core
parameters. As not all parameters can be changed at runtime, only
modifications to parameters that support runtime configuration are
allowed.
The /users/ resource shows all user accounts that can be used with
MaxScale. This resource is further split into two resource collections,
/users/inet and /users/unix, which contain the network users and Linux
accounts respectively.
The /maxscale/ resource now has values for most of the configuration
parameters in the [maxscale] section as well as the version, commit and
uptime information.
The REST API now supports the If-Modified-Since, If-Unmodified-Since,
If-Match and If-None-Match headers and returns the correct response if the
conditional fails.
Added tests for the date parsing and expanded the HTTP header tests in the
REST API test suite.
The resource system now tracks both the time when a resource was last
modified and the revision number of the resource. This allows working
Last-Modified and ETag headers to be generated by the REST API.
The If-Modified-Since and If-None-Match request headers are not yet
processed and using them will always return the resource instead of a 304
Not Modified response.
Returning 204 No Content removes the cost of always sending back the
modified resource. If the modified resource is required, a GET request
should be made to retrieve it.
Updated tests to account for this change.
The listeners under the /services/:service/listeners collection are now
fully JSON API compliant resources.
The listeners could also be exposed as a /listeners collection to easily
group all listener type resources in one place. This approach does has
some semantical and practical problems, namely the fact that each listener
has a many-to-one relationship with its service and listeners by
themselves can't exist alone.
The JSON API specification suggests that the API returns the 403 Forbidden
error when the user does an invalid request. The 400 Bad Request isn't the
ideal error for cases where the syntax is correct but the action being
performed is wrong.
The modules, their types and default values are exposed via the
/maxscale/modules resource. Currently, only a list of resources can be
exposed as the externally exposed module object (MXS_MODULE) does not have
the name and type information in it.
Destroyed servers were still shown as a part of the servers resource
collection.
If a parameter defined in persisted configurations was replaced, the value
would be appended to itself after it was replaced.
Return correct error codes for internal errors.
The server check was checking for old parameter locations.
Service parameters can now be altered with a PUT request to the REST
API. This allows general level parameters to be altered.
Module specific parameters need to be altered with a different mechanism,
namely the module command system. This requires that a generic way to call
a function needs to be devised.
