| Age | Commit message (Collapse) | Author |
|---|
|
This change updates the compute service simulator to use OpenTelemetry
for reporting metrics of the (simulated) hosts as opposed to using
custom event flows.
This approach is more generic, flexible and possibly offers better
performance as we can collect metrics of all services in a single sweep,
as opposed to listening to several services and each invoking the
handlers.
|
|
This change updates the resource model implementation to cache the
remaining work field, which was being computed multiple times during the
same cycle.
|
|
This change removes the event tracer from the repository as we migrate
to the industry standard OpenTelemetry.
|
|
This change removes the event tracer from the OpenDC Workflow service as
we start migrating to the industry standard OpenTelemetry.
|
|
This change integrates the OpenTelemetry Metrics API in the OpenDC
Workflow Service implementation. This replaces the old infrastructure for
gathering metrics.
|
|
This change integrates the OpenTelemetry Metrics API in the OpenDC
Compute Service implementation. This replaces the old infrastructure for
gathering metrics.
|
|
This change removes the dependency on the event tracer in
`opendc-trace`, since we are in the process of migrating OpenDC to
use OpenTelemetry for distributed tracing and metrics.
|
|
This change changes the compute service and users of the compute service
to not rely on the internals of `ComputeServiceImpl` and instead use its
public API.
|
|
This change adds a test suite for the OpenDC compute service.
|
|
This pull request is a preparation for the adding an advanced energy model to OpenDC (#90):
* Add benchmarks for the `opendc-simulator-compute` module, which enables us to quantify the effect on future changes w.r.t. to the energy model on the performance.
* Add `SimResourceTransformer` which can transform resource consumptions. This is useful when transforming from CPU frequency to energy usage for instance.
* Move power models to `opendc-simulator-compute` to be able to be used more generically.
**Breaking API Changes**
* Power models have moved to `opendc-simulator-compute` and now implement the `MachinePowerModel` interface.
|
|
This change moves the power models from the `opendc-compute-simulator`
to the `opendc-simulator-compute` module, since it better fits the scope
of the models and allows them to be re-used for other purposes.
|
|
This change adds support for transforming the resource commands emitted
by the resource consumers. The SimResourceForwarder is modified to also
support transforming the resource commands.
|
|
This change adds benchmarks to the opendc-simulator-compute module in
order to quantify effect of changes on the performance of this module.
|
|
This is the second pull request in the series of pull requests to add a uniform resource consumption model to OpenDC. This pull request focusses on adding dynamic capacity negotiation and propagation between resource consumer and resource provider:
* The generic resource constraint is removed from the interfaces of `opendc-simulator-resources`. Users of the API are expected to use the untyped variants where only the capacity needs to be specified. Users are expected to build higher-level abstractions on top of these interface to represent actual resources (e.g., CPU, disk or network).
* Added benchmarks for the most important implementations of `opendc-simulator-resources`. This allows us to quantify the effects of changes on the runtime.
* The `SimResourceSwitchMaxMin` has been split into a `SimResourceAggregatorMaxMin` and `SimResourceDistributorMaxMin` which respectively aggregate input resources and distribute output resources using max-min fair sharing.
* The `SimResourceConsumer` interface has a new method for receiving capacity change events: `onCapacityChanged(ctx, isThrottled)`
**Breaking API Changes**
* All interfaces in `opendc-simulator-resources`.
|
|
This change re-enables support for VM usage metrics by adding an adapter
for SimResourceConsumer instances that can export the consumer speed.
|
|
This change adds support for dynamically changing the capacity of
resources and propagating this change to consumers.
|
|
This change removes the generic resource constraint (e.g., SimResource)
and replaces it by a simple capacity property. In the future, users
should handle the resource properties on a higher level.
This change simplifies compositions of consumers and providers by not
requiring a translation from resource to capacity.
|
|
This change changes the consumer and context interfaces to expose the
provider capacity and remaining work via the context instance as opposed
to only via the callback. This simplifies aggregation of resources.
|
|
This change adds an initial set of benchmarks for the resource
consumption framework in order to measure the effect of changes on the
performance of the simulator.
|
|
This change re-designs the SimResourceConsumer interface to support in
the future capacity negotiation. This basically means that the consumer
will be informed directly when not enough capacity is available, instead
of after the deadline specified by the consumer.
|
|
This change adds the initial implementation of OpenDC Serverless to the
main repository, which is based on Soufiane Jounaid's work. This commit
introduces the main interfaces into the repository. Later commits will
introduce the other features that Soufiane has implemented previously in
OpenDC.
|
|
This pull request enables aggregate code coverage reports via Jacoco and adds support for coverage tracking via Codecov.
|
|
|
|
This change updates the Github Actions workflow configuration to
generate test reports for the simulator CI builds.
|
|
This change enables Codecov coverage reports when building changes.
|
|
This change will cause the CI to only trigger when opening a pull
request.
|
|
This change updates the dependencies for the OpenDC simulator.
|
|
This change adds support for aggregating code coverage results from the
different modules.
|
|
Bumps [jinja2](https://github.com/pallets/jinja) from 2.11.2 to 2.11.3.
- [Release notes](https://github.com/pallets/jinja/releases)
- [Changelog](https://github.com/pallets/jinja/blob/master/CHANGES.rst)
- [Commits](https://github.com/pallets/jinja/compare/2.11.2...2.11.3)
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
|
|
Bumps [urllib3](https://github.com/urllib3/urllib3) from 1.26.0 to 1.26.3.
- [Release notes](https://github.com/urllib3/urllib3/releases)
- [Changelog](https://github.com/urllib3/urllib3/blob/main/CHANGES.rst)
- [Commits](https://github.com/urllib3/urllib3/compare/1.26.0...1.26.3)
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
|
|
This change implements the CPU energy model with p-states from iCanCloud/E-mc2:
- Only pushed a portion of the code for discussion as not sure if the idea is
on track.
- Inline comments have been added, and formal documents will follow once the
model is finalized.
- The p-state power consumptions are currently hard-coded in a companion
object, which should be improved in the next PR(s).
**Breaking Changes**
- CpuPowerModel: directly interact with the machine it is measuring.
- SimBareMetalMachine: expose the speeds of its CPU cores and its clock
instant.
|
|
This is the first in the series of pull requests to add a uniform resource consumption model to OpenDC. This pull request introduces the `opendc-simulator-resources` module which introduces the primitives with which we can model resource consumption of CPUs, disks and network:
* `SimResourceProvider` represents a provider of some generic resource `R`, which may be consumed via `consume(SimResourceConsumer<R>)`
* `SimResourceConsumer` represents a resource consumers and characterizes how the resource is being consumed.
* `SimResourceSwitch` is a generic scheduler for sharing the capacity of multiple resources across multiple consumers.
- `SimResourceSwitchExclusive`: A space-shared switch - each consumer is allocated a single resource exclusively.
- `SimResourceSwitchMinMax`: A time-shared switch - each consumer gets a fair share of the resource capacity.
* `SimResourceForwarder` converts a consumer in a provider.
**Breaking Changes**
* `ProcessingUnit` and `MemoryUnit` renamed to `SimProcessingUnit` and `SimMemoryUnit` respectively.
* `TimerScheduler` accepts a `CoroutineContext` as opposed to a `CoroutineScope`.
|
|
This change moves the hypervisor implementations to the
opendc-simulator-resources module and makes them generic to the resource
type that is being used (e.g., CPU, disk or networking).
|
|
This change fixes a possible memory leakage issue in TimerScheduler when
a large number of timers was scheduled for the same timestamp.
|
|
This change adds a generic framework for modeling resource consumptions and
adapts opendc-simulator-compute to model machines and VMs on top of
this framework.
This framework anticipates the addition of additional resource types
such as memory, disk and network to the OpenDC codebase.
|
|
This change updates the deployment workflow for opendc.org due to
changes on the server.
|
|
This is the second in the series of pull requests to improve the existing cloud compute model (see #86). This pull request removes the dependency on the bare-metal provisioning code which simplifies experiment setup tremendously:
- Remove bare-metal provisioning code (opendc-metal)
- Remove opendc-core which was a relic of the previous codebase and was only used sparingly.
- Move ownership of Server, Image and Flavor to the compute service. Users are expected to create instances via the compute service.
|
|
|
|
|
|
|
|
|
|
This change removes the unused StateFlow from the utils module. It has
been replaced by the proper implementation in the kotlinx-coroutines
library.
|
|
This change removes the opendc-core module. This module was an artifact
of the old codebase and remained mostly unused. This change removes all
usages of the module and if necessary introduces replacement classes.
|
|
|
|
This change adds the ability to define labels and meta-data for
resources. This can be used in the future to identify servers and pass
data between client and server.
|
|
This is the first of the pull requests in an attempt to improve the existing cloud compute model (see #86). This pull request restructures the compute API and splits the consumer and service interfaces into different modules:
- opendc-compute-api now defines the API interface for the OpenDC Compute module, which can be used by consumers of the OpenDC Compute service.
- opendc-compute-service hosts the service implementation for OpenDC Compute and contains all business logic regarding the IaaS platform (such as scheduling).
- opendc-compute-simulator implements a "compute driver" for the OpenDC Compute platform that simulates submitted workloads.
- Image is now a data-class and does not specify itself the workload to simulate. Instead, the workload should be passed via its tags currently (with key "workload"). In the future, the simulation backend will accept a mapper interface that maps Images to Workloads.
|
|
This change adds more methods for controlling the lifecycle of Server instances.
|
|
This change removes the usage of bare-metal provisioning from the OpenDC
Compute module. This significantly simplifies the experiment setup.
|
|
This change moves the bare-metal provisioning packages outside the
compute module since these modules represent different layers in the
ecosystem and should not be mixed.
|
|
This change introduces the ComputeService interface (previously
VirtProvisioningService) and provides a central implementation in
opendc-compute-service.
Previously, the implementation of this interface was bound to the
simulator package, which meant that independent business logic could not
be re-used without importing the simulator code.
|
