Convert the existing swagger2.0 file to a moden OpenAPI file.
Add missing endpoitns and model definitions.
Signed-off-by: Lucas Roesler <roesler.lucas@gmail.com>
This is not related to the Go version being used, but the min
version a developer can use to contribute.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
Scaling ranges keep the OpenFaaS CE gateway's requests to
within the intended bounds already implemented in other
components.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
OpenFaaS CE exposes metrics about function invocations and
about the gateway itself. OpenFaaS Pro has a richer set of
metrics including HTTP RED.
See also: https://docs.openfaas.com/architecture/metrics/
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
NATS Streaming is deprecated and will be removed from OpenFaaS
CE in a future release for security reasons.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
Sets a new default maximum scale limit of 5 replicas out of
the box for CE users, CE meaning "Community" rather than
"Commercial".
The increment factor of 10 vs 25 should not make a difference
to genuine community and hobbyist users.
Tested and verified with unit tests and hey with a CE cluster
where the maximum limit was reached over several minutes,
finally going back to 1 replica.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
There is no need for OpenFaaS CE to have external auth plugins
since this added extra overhead and was never used.
OpenFaaS Pro retains the option so it can use the OIDC
auth plugin.
It's still possible, as it ever was to put a proxy in front
of any HTTP server like the gateway.
Tested with a local KinD cluster, auth still worked for the
API and UI.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
This change removes the direct functions option which was
used originally for Docker Swarm. The Community Edition will
rely on the faas provider - faas-netes / faasd for load-balancing
of requests.
Direct Functions is required in order to delegate load-balancing
to Istio, Linkerd or some other kind of service mesh.
Tested by deploying a modified gateway image to a KinD cluster,
deploying the env function, and scaling to two replicas. This
balanced the load between the two pods by printing out the names
and then I ran a test with hey which returned 200s for all the
requests.
The prober which was part of the Istio support is no longer
required in the CE gateway so is removed for simplicity.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
NATS Streaming is deprecated and will have no support from
early 2023 by Synadia. Upgrade to OpenFaaS Pro as soon as
possible.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
Return the original upstream response body when the the list request
returns an error. In general, the provider is returning useful and
actionable error messages for the user, the previous code hid this in
the logs and this is easy for user to overlook.
Additionally, remove an early return from error case after fetching
metrics. This looked like a bug and could result in empty api responses
if there was a prometheus error.
Signed-off-by: Lucas Roesler <roesler.lucas@gmail.com>
We now have two write interceptors, with one moved into
faas-provider. This commit makes the gateway use the new
external package and deletes its own.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
During some exploratory testing, I ran into an issue where
the gateway would attempt to scale a deployment from zero
replicas to min, despite there already being min replicas.
Why?
The scaling logic was looking for Available replicas when
it should have looked for Desired replicas. So when a
deployment had zero ready replicas due to readiness checks
failing, the gateway was attempting to scale from zero
to min.
This logic has been corrected and separated from the
a holding pattern where the gateway waits for a ready
replica.
Tested with KinD and an edited function which had a
readiness probe, which was failing and no ready
replicas. As desired, the gateway did not scale to min.
However, when setting desired replicas to zero, the
gateway did scale up as expected.
This change also modifies all print statements for
"seconds" and makes them use 4 decimal places instead of
the default which was a longer, more verbose string for
the logs.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
The welcome message shows the difference between
Pro and CE.
The timeout of 8 seconds was never going to be useful as
a default, so changing to 60 seconds.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
* Removes service min and target metrics from the CE gateway
OpenFaaS Pro metrics are no longer required in OpenFaaS CE
since there is an OpenFaaS Pro gateway available.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
Introduces a single-flight call to a function's health
endpoint to verify that it is registered with an Istio
sidecar (Envoy) before letting the invocation through.
Results are cached for 5 seconds, before a probe is
required again.
Tested without Istio, with probe_functions environment
variable set to true, I saw a probe execute in the logs.
Fixes: #1721 for Istio users.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
When querying for replicas during a scale up event, then the
gateway can overwhelm the provider with requests. This is
especially true under high concurrent load.
The changes in this PR limit the inflight requests.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
The histogram for gateway_functions_seconds excluded the status
code that gives important information for setting up SLOs.
Fixes: #1725
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
* Add service target metric
* Add service min replicas metric
* Add scale type metric
These combined allow new auto-scaling modes and parameters
for OpenFaaS Pro customers.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
The queue type now resides in the provider, so that there is
no risk of a circular reference.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
This is currently being hidden and needs to be bubbled up
into the logs to show an issue with calling list functions
during a deletion in faasd.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
**What**
- Add unit test that verifies the behavior of the logging middleware in
various reponse cases
Signed-off-by: Lucas Roesler <roesler.lucas@gmail.com>
Reduce cache expiry to help support issues where a function
may be scaled down by a user, when it is still in use.
The 5s cache meant that all requests would fail until the
expiry. The 250ms setting is a 20x reduction.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
This appears to have been removed inadvertently whilst
adding scale from zero logic into the gateway.
Going forward, synchronous and asynchronous calls will have a
header populated.
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
