This particular setup had 3 problems:
1. Storing SQL queries as tags is very inefficient as InfluxDB ends up
indexing every query (and they can get pretty large). Storing these
as values instead means we can't always display the SQL as easily.
2. We already instrument ActiveRecord query methods, thus we already
have timing information about database queries.
3. SQL obfuscation is difficult to get right and I'd rather not expose
sensitive data by accident.
While it's useful to keep track of the different versions (Ruby, GitLab,
etc) doing so for every point wastes disk space and possibly also RAM
(which InfluxDB is all to eager to gobble up). If we want to see the
performance differences between different GitLab versions simply looking
at the performance since the last release date should suffice.
This removes the need for Sidekiq and any overhead/problems introduced
by TCP. There are a few things to take into account:
1. When writing data to InfluxDB you may still get an error if the
server becomes unavailable during the write. Because of this we're
catching all exceptions and just ignore them (for now).
2. Writing via UDP apparently requires the timestamp to be in
nanoseconds. Without this data either isn't written properly.
3. Due to the restrictions on UDP buffer sizes we're writing metrics one
by one, instead of writing all of them at once.
This ensures we don't end up wasting resources by tracking method calls
that only take a few microseconds. By default the threshold is 10
milliseconds but this can be changed using the gitlab.yml configuration
file.
InfluxDB escapes double quotes upon output which makes it a pain to deal
with. This ensures that if we're using PostgreSQL we don't store any
queries containing double quotes in InfluxDB, solving the escaping
problem.
When using instrument_methods/instrument_instance_methods we only want
to instrument methods defined directly in a class, not those included
via mixins (e.g. whatever RSpec throws in during development).
In case an externally included method _has_ to be instrumented we can
still use the regular instrument_method/instrument_instance_method
methods.
The methods Instrumentation.instrument_methods and
Instrumentation.instrument_instance_methods can be used to instrument
all methods of a module at once.
The use of ActiveSupport would slow down instrumented method calls by
about 180x due to:
1. ActiveSupport itself not being the fastest thing on the planet
2. caller_locations() having quite some overhead
The use of caller_locations() has been removed because it's not _that_
useful since we already know the full namespace of receivers and the
names of the called methods.
The use of ActiveSupport has been replaced with some custom code that's
generated using eval() (which can be quite a bit faster than using
define_method).
This new setup results in instrumented methods only being about 35-40x
slower (compared to non instrumented methods).
This adds the ability to write application metrics (e.g. SQL timings) to
InfluxDB. These metrics can in turn be visualized using Grafana, or
really anything else that can read from InfluxDB. These metrics can be
used to track application performance over time, between different Ruby
versions, different GitLab versions, etc.
== Transaction Metrics
Currently the following is tracked on a per transaction basis (a
transaction is a Rails request or a single Sidekiq job):
* Timings per query along with the raw (obfuscated) SQL and information
about what file the query originated from.
* Timings per view along with the path of the view and information about
what file triggered the rendering process.
* The duration of a request itself along with the controller/worker
class and method name.
* The duration of any instrumented method calls (more below).
== Sampled Metrics
Certain metrics can't be directly associated with a transaction. For
example, a process' total memory usage is unrelated to any running
transactions. While a transaction can result in the memory usage going
up there's no accurate way to determine what transaction is to blame,
this becomes especially problematic in multi-threaded environments.
To solve this problem there's a separate thread that takes samples at a
fixed interval. This thread (using the class Gitlab::Metrics::Sampler)
currently tracks the following:
* The process' total memory usage.
* The number of file descriptors opened by the process.
* The amount of Ruby objects (using ObjectSpace.count_objects).
* GC statistics such as timings, heap slots, etc.
The default/current interval is 15 seconds, any smaller interval might
put too much pressure on InfluxDB (especially when running dozens of
processes).
== Method Instrumentation
While currently not yet used methods can be instrumented to track how
long they take to run. Unlike the likes of New Relic this doesn't
require modifying the source code (e.g. including modules), it all
happens from the outside. For example, to track `User.by_login` we'd add
the following code somewhere in an initializer:
Gitlab::Metrics::Instrumentation.
instrument_method(User, :by_login)
to instead instrument an instance method:
Gitlab::Metrics::Instrumentation.
instrument_instance_method(User, :save)
Instrumentation for either all public model methods or a few crucial
ones will be added in the near future, I simply haven't gotten to doing
so just yet.
== Configuration
By default metrics are disabled. This means users don't have to bother
setting anything up if they don't want to. Metrics can be enabled by
editing one's gitlab.yml configuration file (see
config/gitlab.yml.example for example settings).
== Writing Data To InfluxDB
Because InfluxDB is still a fairly young product I expect the worse.
Data loss, unexpected reboots, the database not responding, you name it.
Because of this data is _not_ written to InfluxDB directly, instead it's
queued and processed by Sidekiq. This ensures that users won't notice
anything when InfluxDB is giving trouble.
The metrics worker can be started in a standalone manner as following:
bundle exec sidekiq -q metrics
The corresponding class is called MetricsWorker.
Yorick Peterse