1) Only two backends are set up for extra-port
2) The setting is checked to work by connecting directly to servers
3) The server connections are saturated before starting MaxScale
4) MaxScale logs are checked for extra-port-related messages
In the tokenizer we will now recognize the character set names
of MariaDB and return a specific token for those. However, where
a character set name is not expected, it will automatically be
treated as an identifier.
Note that when the character set name is explicitly specified
for a literal string, the name must be prefixed with an underscore.
That is, if the character set name is "latin1", when used when
specifying a literal string, it's used as "_latin1 'a'".
Note that this does not fix the sqlite3 bug causing a leak, but
since the statement will now correctly be parsed, the leak will
not manifest itself.
The mkopcodeh.tcl of sqlite3 version 3110100 has a bug that
manifests itself so that it generates broken code depending on
what keywords there are and in what order. The mkopcodeh.tcl
from 3200000 does not have that problem.
Originally, the sqlite installation was imported into the MaxScale
repository in the one gigantic MaxScale 1.4 -> 2.0 commit.
Consequently, there is no import commit to compare to if you want
to extract all MaxScale specific changes. To make it simpler in the
future, sqlite will now be imported in a commit of its own.
So as to allow the change date to be update at each maintenance
release, we need to have a separate license for each minor release.
In each minor branch there will then be a LICENCE.TXT symbolic link
that points to the actual file.
Generated the new MaxCtrl documentation with the new help output. The help
output now doesn't auto-wrap which causes lines to be longer but it
prevents broken Markdown formatting. It also makes it more readable on
wider terminals.
The number of sessions wasn't always incremented but it was always
decremented. This happened primarily when authentication failed. By making
the management of the counters a part of the object lifecycle, this
problem goes away.
Requiring contiguous buffers removes the need to use mxs::Buffer which
also removes the need to check for buffer boundaries.
Converted all the functions used by get_canonical into `static inline` so
that the compiler knows it can inline them. A few of them weren't `static`
which made the calls to the functions unnecessarily expensive.
The backslash was added instead of assigned. Since the value stored at
that position is always a null byte, assignment and addition would result
in the same outcome.
The help output of all the list and show commands now explains what each
field in the output table means. The generated table will be added at the
end of the help output.
The descriptions table looks like this:
Field Description
----- -----------
Version MaxScale version
Commit MaxScale commit ID
Started At Time when MaxScale was started
Activated At Time when MaxScale left passive mode
Uptime Time MaxScale has been running
Parameters Global MaxScale parameters
The fields are now defined separately and no longer use the object keys as
the names of the values. This makes it clearer as to what the field
definition actually is. Following commits will add a description key into
each object that makes it possible to easily build help output.
Cherry-picked 12e6ada259c38d535003c068a49bc525c33830d8 into 2.3.
Since the maxctrl build process uses add_custom_command, we must
explicitly list the sources that it depends on. With this change, maxctrl
is rebuild whenever the sources change.
The client count was incremented before authentication was complete, and
should be decremented if it fails. Otherwise service connection limit can
be easily reached.
