Connection pooler support
Add support for a connection pooler. The idea is to make it generic enough to
be able to switch between different implementations (e.g. pgbouncer or
odyssey). Operator needs to create a deployment with pooler and a service for
it to access.
For connection pool to work properly, a database needs to be prepared by
operator, namely a separate user have to be created with an access to an
installed lookup function (to fetch credential for other users).
This setups is supposed to be used only by robot/application users. Usually a
connection pool implementation is more CPU bounded, so it makes sense to create
several pods for connection pool with more emphasize on cpu resources. At the
moment there are no special affinity or tolerations assigned to bring those
pods closer to the database. For availability purposes minimal number of
connection pool pods is 2, ideally they have to be distributed between
different nodes/AZ, but it's not enforced in the operator itself. Available
configuration supposed to be ergonomic and in the normal case require minimum
changes to a manifest to enable connection pool. To have more control over the
configuration and functionality on the pool side one can customize the
corresponding docker image.
Co-authored-by: Felix Kunde <felix-kunde@gmx.de>
The current password generation algorithm is extremely deterministic, due to being based on the standard random number generator with a deterministic seed based on the current Unix timestamp (in seconds).
This can lead to a number of security issues, including:
The same passwords being used in different Kubernetes clusters if the operator is deployed in parallel. (This issue was discovered because of four deployments having the same generated passwords due to automatically being deployed in parallel.)
The passwords being easily guessable based on the time the operator pod started when the database was created. (This would typically be present in logs, metrics, etc., that may typically be accessible to more people than should have database access.)
Fix this issue by replacing the current randomness source with crypto/rand, which should produce cryptographically secure random data that is virtually unguessable. This will avoid both of the above problems as each deployment will be guaranteed to have unique, indeterministic passwords.
* add json:omitempty option to ClusterDomain
* Add default value for ClusterDomain
Unfortunately, omitempty in operator configuration CRD doesn't mean that
defauls from operator config object will be picked up automatically.
Make sure that ClusterDomain default is specified, so that even when
someone will set cluster_domain = "", it will be overwritted with a
default value.
Co-authored-by: mlu42 <mlu42pro@gmail.com>
* bump version to 1.4.0 + some polishing
* align version for UI chart
* update user docs to warn for standby replicas
* minor log message changes for RBAC resources
* define postgres-pod clusterrole and align rbac in chart
* align UI chart rbac with operator and update doc
* operator RBAC needs podsecuritypolicy to grant it to postgres-pod
The code added on #818 depends on map sorting to return a static reason
for service annotation changes. To avoid tests flakiness and map sorting
the tests include a `strings.HasPrefix` instead of comparing the whole
string. One of the test cases,
`service_removes_a_custom_annotation,_adds_a_new_one_and_change_another`,
is trying to test the whole reason string.
This commit replaces the test case reason, for only the reason prefix.
It removes the flakiness from the tests. As all the cases (annotation
adding, removing and value changing) are tested before, it's safe to
test only prefixes.
Also, it renames the test name from `TestServiceAnnotations` to
`TestSameService` and introduces a better description in case of test
failure, describing that only prefixes are tested.
The current implementations for `pkg.util.k8sutil.SameService` considers
only service annotations change on the default annotations created by the
operator. Custom annotations are not compared and consequently not
applied after the first service creation.
This commit introduces a complete annotations comparison between the
current service created by the operator and the new one generated based on
the configs. Also, it adds tests on the above-mentioned function.
The [operator parameters][1] already support the
`custom_service_annotations` config.With this parameter is possible to
define custom annotations that will be used on the services created by the
operator. The `custom_service_annotations` as all the other
[operator parameters][1] are defined on the operator level and do not allow
customization on the cluster level. A cluster may require different service
annotations, as for example, set up different cloud load balancers
timeouts, different ingress annotations, and/or enable more customizable
environments.
This commit introduces a new parameter on the cluster level, called
`serviceAnnotations`, responsible for defining custom annotations just for
the services created by the operator to the specifically defined cluster.
It allows a mix of configuration between `custom_service_annotations` and
`serviceAnnotations` where the latest one will have priority. In order to
allow custom service annotations to be used on services without
LoadBalancers (as for example, service mesh services annotations) both
`custom_service_annotations` and `serviceAnnotations` are applied
independently of load-balancing configuration. For retro-compatibility
purposes, `custom_service_annotations` is still under
[Load balancer related options][2]. The two default annotations when using
LoadBalancer services, `external-dns.alpha.kubernetes.io/hostname` and
`service.beta.kubernetes.io/aws-load-balancer-connection-idle-timeout` are
still defined by the operator.
`service.beta.kubernetes.io/aws-load-balancer-connection-idle-timeout` can
be overridden by `custom_service_annotations` or `serviceAnnotations`,
allowing a more customizable environment.
`external-dns.alpha.kubernetes.io/hostname` can not be overridden once
there is no differentiation between custom service annotations for
replicas and masters.
It updates the documentation and creates the necessary unit and e2e
tests to the above-described feature too.
[1]: https://github.com/zalando/postgres-operator/blob/master/docs/reference/operator_parameters.md
[2]: https://github.com/zalando/postgres-operator/blob/master/docs/reference/operator_parameters.md#load-balancer-related-options
* Change error computation on JSON Unmarshall
The [Unmarshall function][1] on the encoding/JSON default library returns
different errors for different go versions. On Go 1.12, the version used
currently on the CI system it returns `json: cannot unmarshal number into
Go struct field PostgresSpec.teamId of type string`. On Go 1.13.5 it
returns `json: cannot unmarshal number into Go struct field
PostgresSpec.spec.teamId of type string`. The new version includes more
details of the whole structure being unmarshelled.
This commit introduces the same error but one level deeper on the JSON
structure. It creates consistency across different Go versions.
[1]: https://godoc.org/encoding/json#Unmarshal
* Create subtests on table test scenarios
The Run method of T allows defining subtests creating hierarchical tests.
It provides better visibility of tests in case of failure. More
details on https://golang.org/pkg/testing/.
This commit converts each test scenario on
pkg/apis/acid.zalan.do/v1/util_test.go to subtests, providing a better
visibility and the debugging environment when working with tests. The
following code snippet shows an error during test execution with
subtests:
```
--- FAIL: TestUnmarshalMaintenanceWindow (0.00s)
--- FAIL: TestUnmarshalMaintenanceWindow/expect_error_as_'From'_is_later_than_'To' (0.00s)
```
It included a `about` field on test scenarios describing the test
purpose and/or it expected output. When a description was provided with
comments it was moved to the about field.
* add validation for PG resources and volume size
* check resource requests also on UPDATE and SYNC + update docs
* if cluster was running don't error on sync
* add CRD manifests with validation
* update documentation
* patroni slots is not an array but a nested hash map
* make deps call tools
* cover validation in docs and export it in crds.go
* add toggle to disable creation of CRD validation and document it
* use templated service account also for CRD-configured helm deployment
* Added possibility to add custom annotations to LoadBalancer service.
* Added parameters for custom endpoint, access and secret key for logical backup.
* Modified dump.sh so it knows how to handle new features. Configurable S3 SSE
For optimization purposes operator was creating a cache map to remember
if service accounts and role binding was deployed to a namespace. This
could lead to a problem, when a namespace was deleted, since this
cache was not synchronized. For the sake of correctness remove the
cache, and check every time if required service account and rbac is
present. In the normal case this introduces an overhead of two API calls
per an event (one to get a service accounts, one to get a role binding),
which should not be a problem, unless proven otherwise.
* And attempt to build with modules and remove glide
* new tools.go file to get code-generator dependency + updated codegen + remove Glide files and update docs
* align config map, operator config, helm chart values and templates
* follow helm chart conventions also in CRD templates
* split up values files and add comments
* avoid yaml confusion in postgres manifests
* bump spilo version and use example for logical_backup_s3_bucket
* add ConfigTarget switch to values
This will set up a continuous wal streaming cluster, by adding the corresponding section in postgres manifest. Instead of having a full-fledged standby cluster as in Patroni, here we use only the wal path of the source cluster and stream from there.
Since, standby cluster is streaming from the master and does not require to create or use databases of it's own. Hence, it bypasses the creation of users or databases.
There is a separate sample manifest added to set up a standby-cluster.
* StatefulSet fsGroup config option to allow non-root spilo
* Allow Postgres CRD to overide SpiloFSGroup of the Operator.
* Document FSGroup of a Pod cannot be changed after creation.
* database.go: substitute hardcoded .svc.cluster.local dns suffix with config parameter
Use the pod's configured dns search path, for clusters where .svc.cluster.local is not correct.
Override clone s3 bucket path
Add possibility to use a custom s3 bucket path for cloning a cluster
from an arbitrary bucket (e.g. from another k8s cluster). For that
a new config options is introduced `s3_wal_path`, that should point
to a location that spilo would understand.
* turns PostgresStatus type into a struct with field PostgresClusterStatus
* setStatus patch target is now /status subresource
* unmarshalling PostgresStatus takes care of previous status field convention
* new simple bool functions status.Running(), status.Creating()
* Config option to allow Spilo container to run non-privileged.
Runs non-privileged by default.
Fixes#395
* add spilo_privileged to manifests/configmap.yaml
* add spilo_privileged to helm chart's values.yaml
Add possibility to mount a tmpfs volume to /dev/shm to avoid issues like
[this](https://github.com/docker-library/postgres/issues/416). To achieve that
two new options were introduced:
* `enableShmVolume` to PostgreSQL manifest, to specify whether or not mount
this volume per database cluster
* `enable_shm_volume` to operator configuration, to specify whether or not mount
per operator.
The first one, `enableShmVolume` takes precedence to allow us to be more flexible.
Client-go provides a https://github.com/kubernetes/code-generator package in order to provide the API to work with CRDs similar to the one available for built-in types, i.e. Pods, Statefulsets and so on.
Use this package to generate deepcopy methods (required for CRDs), instead of using an external deepcopy package; we also generate APIs used to manipulate both Postgres and OperatorConfiguration CRDs, as well as informers and listers for the Postgres CRD, instead of using generic informers and CRD REST API; by using generated code we can get rid of some custom and obscure CRD-related code and use a better API.
All generated code resides in /pkg/generated, with an exception of zz_deepcopy.go in apis/acid.zalan.do/v1
Rename postgres-operator-configuration CRD to OperatorConfiguration, since the former broke naming convention in the code-generator.
Moved Postgresql, PostgresqlList, OperatorConfiguration and OperatorConfigurationList and other types used by them into
Change the type of the Error field in the Postgresql crd to a string, so that client-go could generate a deepcopy for it.
Use generated code to set status of CRD objects as well. Right now this is done with patch, however, Kubernetes 1.11 introduces the /status subresources, allowing us to set the status with
the special updateStatus call in the future. For now, we keep the code that is compatible with earlier versions of Kubernetes.
Rename postgresql.go to database.go and status.go to logs_and_api.go to reflect the purpose of each of those files.
Update client-go dependencies.
Minor reformatting and renaming.
Previously, the operator put pg_hba into the bootstrap/pg_hba key of
Patroni. That had 2 adverse effects:
- pg_hba.conf was shadowed by Spilo default section in the local
postgresql configuration
- when updating pg_hba in the cluster manifest, the updated lines were
not propagated to DCS, since the key was defined in the boostrap
section of Patroni.
Include some minor refactoring, moving methods to unexported when
possible and commenting out usage of md5, so that gosec won't complain.
Per https://github.com/zalando-incubator/postgres-operator/issues/330
Review by @zerg-junior
Assign the list of clusters in the controller with the up-to-date list
of Postgres manifests on Kubernetes during the startup.
Node migration routines launched asynchronously to the cluster
processing rely on an up-to-date list of clusters in the controller to
detect clusters affected by the migration of the node and lock them
when doing migration of master pods. Without the initial list the
operator was subject to race conditions like the one described at
https://github.com/zalando-incubator/postgres-operator/issues/363
Restructure the code to decouple list cluster function required by the
postgresql informer from the one that emits cluster sync events. No
extra work is introduced, since cluster sync already runs in a separate
goroutine (clusterResync).
Introduce explicit initial cluster sync at the end of
acquireInitialListOfClusters instead of relying on an implicit one
coming from list function of the PostgreSQL informer.
Some minor refactoring.
Review by @zerg-junior
Not much changes, except for one function that has been deprecated.
However, unless we find a way to use semantic version comparisons like
'^' on a branch name, we would have to update the apimachinery,
apiextensions-apiserver and code-generator dependencies manually.
Also, slash a linter warning about RoleOriginUnknown being not used.
Run more linters in the gometalinter, i.e. deadcode, megacheck,
nakedret, dup.
More consistent code formatting, remove two dead functions, eliminate
naked a bunch of naked returns, refactor a few functions to avoid code
duplication.
* Allow configuring pod priority globally and per cluster.
Allow to specify pod priority class for all pods managed by the operator,
as well as for those belonging to individual clusters.
Controlled by the pod_priority_class_name operator configuration
parameter and the podPriorityClassName manifest option.
See https://kubernetes.io/docs/concepts/configuration/pod-priority-preemption/#priorityclass
for the explanation on how to define priority classes since Kubernetes 1.8.
Some import order changes are due to go fmt.
Removal of OrphanDependents deprecated field.
Code review by @zerg-junior
There are shortcuts in this code, i.e. we created the deepcopy function
by using the deepcopy package instead of the generated code, that will
be addressed once migrated to client-go v8. Also, some objects,
particularly statefulsets, are still taken from v1beta, this will also
be addressed in further commits once the changes are stabilized.
A repair is a sync scan that acts only on those clusters that indicate
that the last add, update or sync operation on them has failed. It is
supposed to kick in more frequently than the repair scan. The repair
scan still remains to be useful to fix the consequences of external
actions (i.e. someone deletes a postgres-related service by mistake)
unbeknownst to the operator.
The repair scan is controlled by the new repair_period parameter in the
operator configuration. It has to be at least 2 times more frequent than
a sync scan to have any effect (a normal sync scan will update both last
synced and last repaired attributes of the controller, since repair is
just a sync underneath).
A repair scan could be queued for a cluster that is already being synced
if the sync period exceeds the interval between repairs. In that case a
repair event will be discarded once the corresponding worker finds out
that the cluster is not failing anymore.
Review by @zerg-junior
* Improve generting of Scalyr container environment.
Avoid duplicating POD_NAME and POD_NAMESPACE that already bundled
every sidecar.
Do not complain on the lack of SCLALYR_SERVER_HOST, since it is set to
https://upload.eu.scalyr.com in the container we use.
Do not mentioned SCALYR_SERVER_HOST in the error messages, since it is
derived from the cluster name automatically.
Do not show 'persistent volumes are not compatible' errors for the
volumes that failed to be resized because of the other reasons (i.e.
the new size is smaller than the existing one).
* During initial Event processing submit the service account for pods and bind it to a cluster role that allows Patroni to successfully start. The cluster role is assumed to be created by the k8s cluster administrator.
* Up until now, the operator read its own configuration from the
configmap. That has a number of limitations, i.e. when the
configuration value is not a scalar, but a map or a list. We use a
custom code based on github.com/kelseyhightower/envconfig to decode
non-scalar values out of plain text keys, but that breaks when the data
inside the keys contains both YAML-special elememtns (i.e. commas) and
complex quotes, one good example for that is search_path inside
`team_api_role_configuration`. In addition, reliance on the configmap
forced a flag structure on the configuration, making it hard to write
and to read (see
https://github.com/zalando-incubator/postgres-operator/pull/308#issuecomment-395131778).
The changes allow to supply the operator configuration in a proper YAML
file. That required registering a custom CRD to support the operator
configuration and provide an example at
manifests/postgresql-operator-default-configuration.yaml. At the moment,
both old configmap and the new CRD configuration is supported, so no
compatibility issues, however, in the future I'd like to deprecate the
configmap-based configuration altogether. Contrary to the
configmap-based configuration, the CRD one doesn't embed defaults into
the operator code, however, one can use the
manifests/postgresql-operator-default-configuration.yaml as a starting
point in order to build a custom configuration.
Since previously `ReadyWaitInterval` and `ReadyWaitTimeout` parameters
used to create the CRD were taken from the operator configuration, which
is not possible if the configuration itself is stored in the CRD object,
I've added the ability to specify them as environment variables
`CRD_READY_WAIT_INTERVAL` and `CRD_READY_WAIT_TIMEOUT` respectively.
Per review by @zerg-junior and @Jan-M.
* Switchover must wait for the inner goroutine before it returns.
Otherwise, two corner cases may happen:
- waitForPodLabel writes to the podLabelErr channel that has been
already closed by the outer routine
- the outer routine exists and the caller subscribes to the pod
the inner goroutine has already subscribed to, resulting in panic.
The previous commit fe47f9ebea
that touched that code added the cancellation channel, but didn't bother
to actually wait for the goroutine to be cancelled.
Per report and review from @valer-cara.
Original issue: https://github.com/zalando-incubator/postgres-operator/issues/342
The old way of specifying it with the annotation is deprecated and not
available in recent Kubernetes versions. We will keep it there anyway
until upgrading to the new go-client that is incompatible with those
versions.
Per report from @schmitch
* Define sidecars in the operator configuration.
Right now only the name and the docker image can be defined, but with
the help of the pod_environment_configmap parameter arbitrary
environment variables can be passed to the sidecars.
* Refactoring around generatePodTemplate.
Original implementation of per-cluster sidecars by @theRealWardo
Per review by @zerg-junior and @Jan-M
Call Patroni API /config in order to set special options that are
ignored when set in the configuration file, such as max_connections.
Per https://github.com/zalando-incubator/postgres-operator/issues/297
* Some minor refacoring:
Rename Cluster ManualFailover to Swithover
Rename Patroni Failover to Switchover
Add more details to error messages and comments introduced in this PR.
Review by @zerg-junior
After an unsuccessful initial cluster sync it may happen that the
cluster statefulset is empty. This has been made more likely since
88d6a7be3, since it has introduced syncing volumes before statefulsets,
and the volume sync mail fail for different reasons (i.e. the volume has
been shrinked, or too many calls to Amazon).
Some special patroni postgresql parameters, like max_connections,
should reside in the bootstrap.dcs.postgresql.parameters section
to come into effect.
When there is an error happening upon deletion of the Kubernetes object
belonging to the cluster being removed, it makes no sense to abort the
deletion: the manifest will be removed anyway, therefore all the objects
after the one we aborted at will stay forever.
Do not use statefulset number of pods to figure out running ones
for volume resizing, since the statefulset pointer could be nil.
Instead, look at the actual running pods.
ReSearchITEng
Felix Kunde
Dmitry Dolgov
Fredrik Østrem
zimbatm
grantlanglois
Jonathan Juares Beber
Vito Botta
Armin Nesiren
Thomas Runyon
Eric
Weilu Jia
Sergey Dudoladov
Rafia Sabih
Markus
Maxim Ivanov
Erik Inge Bolsø
teuto.net Netzdienste GmbH
Aaron Miller
Stephane T
Vineeth Reddy
Rafał Kupka
Jan Mussler
Arve Knudsen
Noah Kantrowitz
Oleksii Kliukin
Oleksii Kliukin
Sergey Dudoladov
Manuel Gómez
Oleksii Kliukin