By allowing transactions to the master to end even if the server is in
maintenance mode makes it possible to terminate connections at a known
point. This helps prevent interrupted transactions which can help reduce
errors that are visible to the clients.
If a server with zero weight was chosen as the only candidate, it was
possible that the starting minimum value was smaller than the server
score. This would mean that a candidate wouldn't be chosen if the score
was too high. To preven this, the values are capped to a value smaller
than the initial minimum score.
Increasing counter sizes from int to long for averages.
Rename random functions to end with _co instead of _exclusive to
indicate range [close, open[, and to allow future suffixes oc, cc and oo.
When a readwritesplit session has a connection to a master server, servers
of the same rank as the master are used. If no master connection is
available, the server with the highest rank among all connected servers is
used. If there are no open connections, the server with the best rank is
chosen and a connection to it is made.
Connections with different rank values than what is the current rank value
of the session will be discarded. This reduces the use of server with
different ranks when the master server of a session fails. Without the
active pruning of connections, slave connections to primary clusters
without masters would remain in use even after the primary master
fails. This guarantees full switchover to a secondary cluster if a master
change occurs.
Although the default value is the maximum value of a signed 32-bit
integer, the value is stored as a 64-bit integer. The integer type
conversion functions return 64-bit values so storing it as one makes
sense.
Currently values higher than the default are allowed but the accepted
range of input should be restricted in the future.
Readwritesplit now respects server ranks. When servers are selected for
either routing or connection creation, the servers are partitioned by
their rank into sets of servers. These sets of servers are never mixed so
the end result is that only servers of the same rank are considered for
candidacy.
The master selection is slightly different: the server with the best rank
that is capable of acting as a master is chosen. This means that a session
can have a master with a lower rank and slaves with higher ranks than the
master. In most cases this actually is the preferred behavior as the rank
is used to prioritize usage but not outright prevent it.
The connection creation is now internal to RWSplitSession. This makes the
code more readable by removing the need to pass parameters and allowing
easier reuse of existing functions. The various conditions require to
create connections are now also checked in only one place.
Readwritesplit now picks the best available master if no open master
connection is available. This is required if the server rank is to be
taken into account when master selection is done.
The candidate selection code used default values that would cause reads
past buffers. The code could also dereference the end iterator which
causes undefined behavior.
The lazy connection creation reduces the burden that short sessions place
on the backend servers. This also prevents the problems caused by early
disconnections that happen when only one server is used but multiple
connections are created. This does not solve the problem (MXS-619) but it
does mitigate it to acceptable levels.
This commit also adds a change to the weighting algorithm that prefers
existing connections over unopened ones. This helps avoid the
flip-flopping that happens when the absolute scores are very similar. The
hard-coded value might need to be tuned once testing is done.
Given the fact that there exist only three possible categories, the map
can be replaced with a static array that needs no memory
allocations. Making this array thread-local allows it to be reused which
places an upper limit on the number of memory allocations.
If a master is found but it is being drained, the connection attempt
is rejected if the master failure mode is fail_instantly.
In that case the logged message makes it plain that it is the draining
that is the reason for the connection attempt to fail.
Most of the ones still remaining outside are special cases.
Also, removed locking from status manipulation functions as it
has not been required for quite some time.
For lifetime management keep RWBackends in a vector of unique_ptrs.
RWSplitSession keeps the unique_ptrs very private, and provides a vector
of plain pointers for all other interfaces.
This is essentially just a search and replace to change SRWBackend to
RWBackend* and SRWBackendList to PRWBackends, a vector of a raw
pointers. In the next few commits vector<unique_ptr<RWBackend>>
will be used for life time management.
There are a lot of diffs from the global search and replace. Only a few manual
edits had to be done.
list-src -x build | xargs sed -ri 's/SRWBackends/prwbackends/g'
list-src -x build | xargs sed -ri 's/const mxs::SRWBackend\&/const mxs::RWBackend\*/g'
list-src -x build | xargs sed -ri 's/const SRWBackend\&/const RWBackend\*/g'
list-src -x build | xargs sed -ri 's/mxs::SRWBackend\&/mxs::RWBackend\*/g'
list-src -x build | xargs sed -ri 's/mxs::SRWBackend/mxs::RWBackend\*/g'
list-src -x build | xargs sed -ri 's/SRWBackend\(\)/nullptr/g'
list-src -x build | xargs sed -ri 's/mxs::SRWBackend\&/mxs::RWBackend\*/g'
list-src -x build | xargs sed -ri 's/mxs::SRWBackend/mxs::RWBackend\*/g'
list-src -x build | xargs sed -ri 's/SRWBackend\&/RWBackend\*/g'
list-src -x build | xargs sed -ri 's/SRWBackend\b/RWBackend\*/g'
list-src -x build | xargs sed -ri 's/prwbackends/PRWBackends/g'
The servers with a zero weight would be always used over ones that have a
weight. This means that the behavior was inverted and caused the
mxs2054_hybrid_cluster test to fail in 2.3.
Also fixed a typo in the deprecation message.
By biasing the values of all counter type scores to positive integers, the
server weights are always taken into use.
This fixes the case when weights were ignored until all score base values
were larger than zero (the mxs922_server test).
The additions into the server.h header used C++ language which caused C
programs to fail to compile. Moved the implementation of the EMAverage
class into the private Server class in the server.hh header and exposed it
via functions in the server.h header. Also temporarily moved
almost_equal_server_scores into the public server.hh as there is no
service.hh header.
See script directory for method. The script to run in the top level
MaxScale directory is called maxscale-uncrustify.sh, which uses
another script, list-src, from the same directory (so you need to set
your PATH). The uncrustify version was 0.66.
The math becomes simpler when the weight is inverted, i.e. a simple multiplication
to get the (inverse) score. Inverse weights are normalized to the range [0..1] where a lower
number is a higher weight,
The enum select_criteria_t is used to provide a std::function that takes the backends
as vector (rather than the prior pairwise compares) and returns the best backend.
This commit refactors slave selection. The compare is still pair-wise but isolated into a small run_comparison() function. The function get_slave_candidate() is used when new connections are created, which I both moved and modified (had to move due to scoping), so diff is off.
The slave selection for routing: get_slave_backend() now has the filtering logic from old get_slave_backend() and compare_backends(), the latter of which is removed.
Backend functions mostly take shared_ptr<SRWBackend> in various forms (as is, const ref, in a container). Ideally the shared_ptr would be used only to where it is really needed, and either naked ptrs or references to RWBackend would be used. This refactor does not address that issue, but compounds it by using even deeper shared_ptr structures. Fixing that in a future commit.
The main piece of code, slave selection (backend_cmp_response_time), uses the available
method of pair-wise comparison of slaves. This will be changed to selection using all
available slaves, along with removal of hard coded values.
The configuration doesn't need to be contained in shared pointer as each
session holds its own version of it. This removes most of the overhead in
configuration reloading. The only thing that's left is any overhead added
by the use of thread-local storage.
By using a shared pointer instead of a plain object, we can replace the
router configuration without it affecting existing sessions. This is a
change that is required to enable runtime reconfiguration of
readwritesplit.
Allowing calls to select_connect_backend_servers even when all slaves are
connected solves the debug assertion in select_connect_backend_servers
that happens when the execution of a queued query causes a new connection
to be created.
The logging in the RWSplit::select_connect_backend_servers was quite
cumbersome and doing the logging when the connection is created makes for
a more information rich output (we know what servers were taken into use
and when).
Changed the log messages to use the names of the servers instead of the
address and port.
