In-Network Computing for App-Centric Micro-Services
draft-sarathchandra-coin-appcentres-03
The information below is for an old version of the document.
| Document | Type |
This is an older version of an Internet-Draft whose latest revision state is "Expired".
|
|
|---|---|---|---|
| Authors | Dirk Trossen , Chathura Sarathchandra , Michael Boniface | ||
| Last updated | 2020-10-23 | ||
| RFC stream | (None) | ||
| Formats | |||
| Stream | Stream state | (No stream defined) | |
| Consensus boilerplate | Unknown | ||
| RFC Editor Note | (None) | ||
| IESG | IESG state | I-D Exists | |
| Telechat date | (None) | ||
| Responsible AD | (None) | ||
| Send notices to | (None) |
draft-sarathchandra-coin-appcentres-03
COIN D. Trossen
INTERNET-DRAFT Huawei
Intended Status: Informational C. Sarathchandra
Expires: April 25, 2021 InterDigital Inc.
M. Boniface
University of Southampton
October 23, 2020
In-Network Computing for App-Centric Micro-Services
draft-sarathchandra-coin-appcentres-03
Abstract
The application-centric deployment of 'Internet' services has
increased over the past ten years with many millions of applications
providing user-centric services, executed on increasingly more
powerful smartphones that are supported by Internet-based cloud
services in distributed data centres, the latter mainly provided by
large scale players such as Google, Amazon and alike. This draft
outlines a vision for evolving those data centres towards executing
app-centric micro-services; we dub this evolved data centre as an
AppCentre. Complemented with the proliferation of such AppCentres at
the edge of the network, they will allow for such micro-services to
be distributed across many places of execution, including mobile
terminals themselves, while specific micro-service chains equal
today's applications in existing smartphones. We outline the key
enabling technologies that needs to be provided for such evolution to
be realized, including references to ongoing IETF work in some
areas.
Status of this Memo
This Internet-Draft is submitted to IETF in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as
Internet-Drafts.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
Trossen, et al. Expires April 25, 2021 [Page 1]
INTERNET DRAFT App-Centric Micro-Services
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/1id-abstracts.html
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html
Copyright and License Notice
Copyright (c) 2019 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License.
Table of Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3 Use Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.1 Mobile Application Function Offloading . . . . . . . . . . 5
3.2 Interactive Real-time Applications . . . . . . . . . . . . 7
3.3 Distributed AI . . . . . . . . . . . . . . . . . . . . . . 7
3.4 Content Delivery Networks . . . . . . . . . . . . . . . . . 8
3.5 Compute-Fabric-as-a-Service (CFaaS) . . . . . . . . . . . . 8
4 Requirements Derived from Use Cases . . . . . . . . . . . . . . 9
5 Enabling Technologies . . . . . . . . . . . . . . . . . . . . . 10
5.1 Application Packaging . . . . . . . . . . . . . . . . . . . 10
5.2 Service Deployment . . . . . . . . . . . . . . . . . . . . 11
5.3 Compute Inter-Connection at Layer 2 . . . . . . . . . . . . 12
5.4 Service Routing . . . . . . . . . . . . . . . . . . . . . . 13
5.5 Constraint-based Forwarding Decisions . . . . . . . . . . . 14
5.6 Collective Communication . . . . . . . . . . . . . . . . . 14
5.7 State Synchronization . . . . . . . . . . . . . . . . . . . 15
5.8 Dynamic Contracts . . . . . . . . . . . . . . . . . . . . . 15
6 Security Considerations . . . . . . . . . . . . . . . . . . . . 15
7 IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 15
8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 15
9 References . . . . . . . . . . . . . . . . . . . . . . . . . . 16
9.1 Normative References . . . . . . . . . . . . . . . . . . . 16
Trossen, et al. Expires April 25, 2021 [Page 2]
INTERNET DRAFT App-Centric Micro-Services
9.2 Informative References . . . . . . . . . . . . . . . . . . 16
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 18
Trossen, et al. Expires April 25, 2021 [Page 3]
INTERNET DRAFT App-Centric Micro-Services
1 Introduction
With the increasing dominance of smartphones and application markets,
the end-user experiences today have been increasingly centered around
the applications and the ecosystems that smartphone platforms create.
The experience of the 'Internet' has changed from 'accessing a web
site through a web browser' to 'installing and running an application
on a smartphone'. This app-centric model has changed the way services
are being delivered not only for end-users, but also for business-to-
consumer (B2C) and business-to-business (B2B) relationships.
Designing and engineering applications is largely done statically at
design time, such that achieving significant performance improvements
thereafter has become a challenge (especially, at runtime in response
to changing demands and resources). Applications today come
prepackaged putting them at disadvantage for improving efficiency due
to the monolithic nature of the application packaging. Decomposing
application functions into micro-services [MSERVICE1] [MSERVICE2]
allows applications to be packaged dynamically at run-time taking
varying application requirements and constraints into consideration.
Interpreting an application as a chain of micro-services, allows the
application structure, functionality, and performance to be adapted
dynamically at runtime in consideration of tradeoffs between quality
of experience, quality of service and cost.
Interpreting any resource rich networked computing (and storage)
capability not just as a pico or micro-data centre, but as an
application-centric execution data centre (AppCentre), allows
distributed execution of micro-services. Here, the notion of an
'application' constitutes a set of objectives being realized in a
combined packaging of micro-services under the governance of the
'application provider'. These micro-services may then be deployed on
the most appropriate AppCentre (edge/fog/cloud resources) to satisfy
requirements under varying constraints. In addition, the high degree
of distribution of application and data partitions, and compute
resources offered by the execution environment decentralizes control
between multiple cooperating parties (multi-technology, multi-domain,
multi-ownership environments). Furthermore, compute resource
availability may be volatile, particularly when moving along the
spectrum from well-connected cloud resources over edge data centres
to user-provided compute resources, such as (mobile) terminals or
home-based resources such as NAS and IoT devices.
We believe that the emergence of AppCentreS will democratize
infrastructure and service provision to anyone with compute resources
with the notion of applications providing an element of governing the
execution of micro-services. This increased distribution will lead to
new forms of application interactions and user experiences based on
Trossen, et al. Expires April 25, 2021 [Page 4]
INTERNET DRAFT App-Centric Micro-Services
cooperative AppCentreS (pico-micro and large cloud data centres), in
which applications are being designed, dynamically composed and
executed.
2 Terminology
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in RFC 2119 [RFC2119].
3 Use Cases
Although our motivation for the 'AppCentre' term stems from the
(mobile) application ecosystem, the use cases in this section are not
limited to mobile applications only. Instead, we interpret
'applications' as a governing concept of executing a set of micro-
services where the 'application provider' can reach from those
realizing mobile applications over novel network applications to
emerging infrastructure offerings serving a wide range of
applications in a purpose- (and therefore application-)agnostic
manner. The following use cases provide examples for said spectrum of
applications.
3.1 Mobile Application Function Offloading
Partitioning an application into micro-services allows for denoting
the application as a collection of functions for a flexible
composition and a distributed execution, e.g., most functions of a
mobile application can be categorized into any of three, "receiving",
"processing" and "displaying" function groups.
Any device may realize one or more of the micro-services of an
application and expose them to the execution environment. When the
micro-service sequence is executed on a single device, the outcome is
what you see today as applications running on mobile devices.
However, the execution of functions may be moved to other (e.g., more
suitable) devices which have exposed the corresponding micro-services
to the environment. The result of the latter is flexible mobile
function offloading, for possible reduction of power consumption
(e.g., offloading CPU intensive process functions to a remote server)
or for improved end user experience (e.g., moving display functions
to a nearby smart TV).
The above scenario can be exemplified in an immersive gaming
application, where a single user plays a game using a VR headset. The
headset hosts functions that "display" frames to the user, as well as
the functions for VR content processing and frame rendering combining
with input data received from sensors in the VR headset. Once this
Trossen, et al. Expires April 25, 2021 [Page 5]
INTERNET DRAFT App-Centric Micro-Services
application is partitioned into micro-services and deployed in an
app-centric execution environment, only the "display" micro-service
is left in the headset, while the compute intensive real-time VR
content processing micro-services can be offloaded to a nearby
resource rich home PC, for a better execution (faster and possibly
higher resolution generation).
Figure 1 shows one realization of the above scenario, where a 'DPR
app' is running on a mobile device (containing the partitioned
Display(D), Process(P) and Receive(R) micro services) over an SDN
network. The packaged applications are made available through a
localized 'playstore server'. The application installation is
realized as a 'service deployment' process (Section 5.2.), combining
the local app installation with a distributed micro-service
deployment (and orchestration) on most suitable AppCentreS
('processing server').
+----------+ Processing Server
Mobile | +------+ |
+---------+ | | P | |
| App | | +------+ |
| +-----+ | | +------+ |
| |D|P|R| | | | SR | |
| +-----+ | | +------+ | Internet
| +-----+ | +----------+ /
| | SR | | | /
| +-----+ | +----------+ +------+
+---------+ /|SDN Switch|_____|Border|
\ +-------+ / +----------+ | SR |
\| 5GAN |/ | +------+
+-------+ |
+---------+ |
|+-------+| +----------+
||Display|| /|SDN Switch|
|+-------+| +-------+ / +----------+
|+-------+| /|WIFI AP|/ \
|| D || / +-------+ +--+
|+-------+|/ |SR|
|+-------+| /+--+
|| SR || +---------+
|+-------+| |Playstore|
+---------+ | Server |
TV +---------+
Figure 1: Application Function Offloading Example
Such localized deployment could, for instance, be provided by a
visiting site, such as a hotel or a theme park. Once the 'processing'
Trossen, et al. Expires April 25, 2021 [Page 6]
INTERNET DRAFT App-Centric Micro-Services
micro-service is terminated on the mobile device, the 'service
routing' (SR) elements in the network (Section 5.4.) routes requests
to the previously deployed 'processing' micro-service running on the
'processing server' AppCentre over an existing SDN network. As an
extension to the above scenarios, we can also envision that content
from one processing micro-service may be distributed to more than one
display micro-service, e.g., for multi/many-viewing scenarios.
3.2 Interactive Real-time Applications
There has been a recent shift from applications that provide single-
user experiences, such as the ones described in the previous section
to collaborative/cooperative experiences that are highly interactive
with strict real-time requirements, such as collaborative working,
education, multi-user gaming, and mixed/virtual reality. This leads
to increasing interaction where input (e.g., gesture, gaze, touch,
movement) and output (e.g., visual display, sound, and actuation)
needs to be processed within strict timing constraints and
synchronized to ensure temporal and spatial consistency with local
and distant users. App-centric design allows functions with high data
and process coupling to be modularized, deployed and executed, such
that the subset of micro-services is cooperatively executed towards
optimizing the interactive experiences.
The same example of the previous section can be envisaged in a multi-
player gaming scenario. Here the micro-services that need to be
executed cooperatively are executed in a localized and synchronized
manner to ensure player coordination and synchronized state between
collaborating players.
3.3 Distributed AI
There is a growing range of use cases demanding for the realization
of AI capabilities among distributed endpoints. Such demand may be
driven by the need to increase overall computational power for large-
scale problems. Other solutions may desire the localization of
reasoning logic, e.g., for deriving attributes that better preserve
privacy of the utilized raw input data. Examples for large-scale AI
problems include biotechnology and astronomy related reasoning over
massive amounts of observational input data. Examples for localizing
input data for privacy reasons include radar-like application for the
development of topological mapping data based on (distributed) radio
measurements at base stations (and possibly end devices), while the
processing within radio access networks (RAN) already constitute a
distributed AI problem to a certain extent albeit with little
flexibility in distributing the execution of the AI logic.
Reasoning frameworks, such as TensorFlow, may be utilized for the
Trossen, et al. Expires April 25, 2021 [Page 7]
INTERNET DRAFT App-Centric Micro-Services
realization of the (distributed) AI logic, building on remote service
invocation through protocols such as gRPC [GRPC] or MPI [MPI] with
the intention of providing an on-chip NPU (neural processor unit)
like abstraction to the AI framework.
3.4 Content Delivery Networks
Delivery of content to end users often relies on Content Delivery
Networks (CDNs) storing said content closer to end users for latency
reduced delivery with DNS-based indirection being utilized to serve
the request on behalf of the origin server. From the perspective of
this draft, a CDN can be interpreted as a (network service level)
application with distributed logic for distributing content from the
origin server to the CDN ingress and further to the CDN replication
points which ultimately serve the user-facing content requests.
Studies such as those in [FCDN] have shown that content distribution
at the level of named content, utilizing efficient (e.g., Layer 2)
multicast for replication towards edge CDN nodes, can significantly
increase the overall network and server efficiency. It also reduces
indirection latency for content retrieval as well as reduces required
edge storage capacity by benefiting from the increased network
efficiency to renew edge content more quickly against changing
demand.
3.5 Compute-Fabric-as-a-Service (CFaaS)
App-centric execution environments, consisting of Layer 2 connected
AppcentreS, provide the opportunity for infrastructure providers to
offer CFaaS type of offerings to application providers. Those app
providers utilize the compute fabric exposed by this CFaaS offering
for the purposes defined through their applications. In other words,
the compute resources can be utilized to execute the desired micro-
services of which the application is composed, while utilizing the
inter-connection between those compute resources to do so in a
distributed manner.
We foresee those CFaaS offerings to be tenant-specific, a tenant here
defined as the provider of at least one application. For this, we
foresee an interaction between CFaaS provider and tenant to
dynamically select the appropriate resources to define the demand
side of the fabric. Conversely, we also foresee the supply side of
the fabric to be highly dynamic with resources being offered to the
fabric through, e.g., user-provided resources (whose supply might
depend on highly context-specific supply policies) or infrastructure
resources of intermittent availability such as those provided through
road-side infrastructure in vehicular scenarios. The resulting
dynamic demand-supply matching establishes a dynamic nature of the
Trossen, et al. Expires April 25, 2021 [Page 8]
INTERNET DRAFT App-Centric Micro-Services
compute fabric that in turn requires trust relationships to be built
dynamically between the resource provider(s) and the CFaaS provider.
This also requires the communication resources to be dynamically
adjusted to interconnect all resources suitably into the (tenant-
specific) fabric exposed as CFaaS.
4 Requirements Derived from Use Cases
The following requirements are derived from the presented use cases
in Section 3.1. to 3.5., numbered according to the use case numbers
(as main item numbers) although those requirements apply in some
cases across more than one of the presented use cases.
Req 1.1: Any app-centric execution environment MUST provide means for
routing of service requests between resources in the
distributed environment.
Req 1.2: Any app-centric execution environment MUST provide means for
dynamically choosing the best possible micro-service
sequence (i.e., chaining of micro-services) for a given
application experience. Means for discovering suitable
micro-service SHOULD be provided.
Req 1.3: Any app-centric execution environment MUST provide means for
pinning the excution of a specific micro-service to a
specific resource instance in the distributed environment.
Req 1.4: Any app-centric execution environment SHOULD provide means
for packaging micro-services for deployments in distributed
networked computing environments. The packaging MAY include
any constraints regarding the deployment of service
instances in specific network locations or compute
resources. Such packaging SHOULD conform to existing
application deployment models, such as mobile application
packaging, TOSCA orchestration templates or tar balls or
combinations thereof.
Req 2.1: Any app-centric execution environment MUST provide means for
real-time synchronization and consistency of distributed
application states.
Req 3.1: Any app-centric execution environment MUST provide means to
specify the constraints for placing (AI) execution logic in
certain logical execution points (and their associated
physical locations).
Req 3.2: Any app-centric execution environment MUST provide support
for app/micro-service specific invocation protocols.
Trossen, et al. Expires April 25, 2021 [Page 9]
INTERNET DRAFT App-Centric Micro-Services
Req 4.1: Any app-centric execution environment SHOULD utilize Layer 2
multicast transmission capabilities for responses to
concurrent service requests.
Req 5.1: Any app-specific execution environment SHOULD expose means
to specify the requirements for the tenant-specific compute
fabric being utilized for the app execution.
Req 5.2: Any app-specific execution environment SHOULD allow for
dynamic integration of compute resources into the compute
fabric being utilized for the app execution; those resources
include, but are not limited to, end user provided
resources.
Req 5.3: Any app-specific execution environment MUST provide means to
optimize the inter-connection of compute resources,
including those dynamically added and removed during the
provisioning of the tenant-specific compute fabric.
Req 5.4: Any app-specific execution environment MUST provide means
for ensuring availability and usage of resources is
accounted for.
5 Enabling Technologies
This section discusses a number of enabling technologies relevant for
the realization of the app-centric micro-service vision laid out in
this draft.
EDITOR NOTE: Section 5 will be updated to include the addressing of
specific requirements listed in Section 4.
5.1 Application Packaging
Applications often consist of one or more sub-elements (e.g., audio,
visual, hepatic elements) which are 'packaged' together, resulting in
the final installable software artifact. Conventionally, application
developers perform the packaging process at design time, by packaging
a set of software components as a (often single) monolithic software
package, for satisfying a set of predefined application
requirements.
Decomposing micro-services of an application, and then executing them
on peer execution points in AppCentreS (e.g., on an app-centric
serverless runtime [SRVLESS]) can be done with design-time planning.
Micro-service decomposition process involves, defining clear
boundaries of the micro-service (e.g., using wrapper classes for
Trossen, et al. Expires April 25, 2021 [Page 10]
INTERNET DRAFT App-Centric Micro-Services
handling input/output requests), which could be done by the
application developer at design-time (e.g., through Android app
packaging by including, as part of the asset directory, a service
orchestration template [TOSCA] that describes the decomposed micro-
services). Likewise, the peer execution points could be 'known' to
the application (e.g., using well-known and fixed peer execution
points on AppCentreS) and incorporated with the micro-services by the
developer at design-time.
Existing programming frameworks address decomposition and execution
of applications centering around other aspects such as concurrency
[ERLANG]. For decomposing at runtime, application elements can be
profiled using various techniques such as dynamic program analysis or
dwarf application benchmarks. The local profiler information can be
combined with the profiler information of other devices in the
network for improved accuracy. The output of such a profiler process
can then be used to identify smaller constituting sub-components of
the application in forms of pico-services, their interdependencies
and data flow (e.g., using caller/callee information, instruction
usage). Due to the complex nature of resulting application structure
and therefore its increased overhead, in most cases, it may not be
optimal to decompose applications at the pico level. Therefore, one
may cluster pico-services into micro-services with common
characteristics, enabling a meaningful (e.g., clustering pico-
services with same resource dependency) and a performant
decomposition of applications. Characteristics of micro-services can
be defined as a set of concepts using an ontology language, which can
then be used for clustering similar pico-services into micro-
services. Micro-services may then be partitioned along their
identified borders. Moreover, mechanisms for governance, discovery
and offloading can be employed for 'unknown' peer execution points on
AppCentreS with distributed loci of control.
Therefore, with this app-centric model, application packaging can be
done at runtime by constructing micro-service chains for satisfying
requirements of experiences (e.g., interaction requirements), under
varying constraints (e.g., temporal consistency between multiple
players within a shared AR/VR world)[SCOMPOSE]. Such packaging
includes mechanisms for selecting the best possible micro-services
for a given experience at runtime in the multi-technology
environment. These run-time packaging operations may continuously
discover the 'unknown' and adapt towards an optimal experience. Such
decision mechanisms handle the variability, volatility and scarcity
within this multi-X framework.
5.2 Service Deployment
The service function chains, constituting each individual
Trossen, et al. Expires April 25, 2021 [Page 11]
INTERNET DRAFT App-Centric Micro-Services
application, will need deployment mechanisms in a true multi-X
(multi-user, multi-infrastructure, multi-domain) environment
[SDEPLOY1][SDEPLOY2]. Most importantly, application installation and
orchestration processes are married into one, as a set of procedures
governed by device owners directly or with delegated authority.
However, apart from extending towards multi-X environments, the
process also needs to cater for changes in the environment, caused,
e.g., by movement of users, new pervasive sensors/actuators, and
changes to available infrastructure resources. Methods are needed to
deploy service functions as executable code into chosen service
execution points. Those methods need to support the various endpoint
(e.g., device stacks, COTS stacks, etc.) and service function
realizations, e.g., through utilizing existing and emerging
virtualization techniques.
A combination of application installation procedure and orchestrated
service deployment can be achieved by utilizing the application
packaging with integrated service deployment templates described in
Section 5.1 such that the application installation procedure on the
installing device is being extended to not only install the local
application package but also extract the service deployment template
for orchestrating with the localized infrastructure, using, for
instance, REST APIs for submitting the template to the orchestrator.
5.3 Compute Inter-Connection at Layer 2
While Layer2 switching technologies have long proliferated in data
centre deployments, recent developments have advanced the
capabilities for interconnecting distributed computing resources over
Layer2 technologies. For instance, the efforts in 3GPP on so-called
'5G LAN' (or Vertical LAN) [SA2-5GLAN] allow for establishing a
Layer2 bearer between participating compute entities, using a
vertical-specific LAN identifier for packet forwarding between the
distributed Layer2 entities. Combined with Layer2 technology in data
centres as well as office and home networks alike, this enables the
deployment of services in vertical (Layer2) networks, interconnecting
with other Internet-based services through dedicated service
gateways.
Real deployments and realizations will have to show the scalability
of this approach but it points into a direction where application or
service-specific deployments could potentially 'live' entirely in
their own vertical network, interconnecting only based on need (which
for many services may not exist). From the application's or service's
perspective, the available compute resource pool will look no
different from that being realized in a single data centre albeit
with the possibility to being highly distributed now among many
(e.g., edge) data centres as well as mobile devices.
Trossen, et al. Expires April 25, 2021 [Page 12]
INTERNET DRAFT App-Centric Micro-Services
In such a deployment, it is interesting to study the realization of
suitable service routing capabilities, leading us to the next
technology area of interest.
5.4 Service Routing
Routing service requests is a key aspect within a combined compute
and network infrastructure in order to enable true end-to-end
experiences across distributed application execution points
provisioned on distant cloud, edge and device-centric resources. Once
the micro-services are packaged and deployed in such highly
distributed micro-data centres, the routing mechanisms must ensure
efficient information exchange between corresponding micro-services,
e.g., at the level of service requests, within the multi-technology
execution environment.
Routing here becomes a problem of routing micro-service requests, not
just packets, as done through IP. This calls for some form of 'flow
affinity' that allows for treating several packets as part of a
request semantic. This is important, e.g., for mobility (avoiding to
send some packets of a larger request to one entity, while other
packets are sent to another one, therefore creating incomplete
information at both entities as a result). Also, when applying
constraints to the forwarding of packets (discussed in more detail in
Section 5.6), it is important to apply the actions across the packets
of the request rather than individually.
Another key aspect is that of addressing services. Traditionally, the
combination of the Domain Naming Service (DNS) and IP routing has
been used for this purpose. However, the advent of virtualization
with use cases such as those outlined in Section 3 (such as those on
app-specific micro-services on mobile devices) have made it
challenging to further rely on the DNS. Apart from the initial delay
observed when resolving a service name into a locator for the first
time, the long delay in updating DNS entries to 'point' to the right
micro-service instances prohibits offloading to dynamically created
service instances. If one was to use the DNS, one would be updating
the DNS entries at a high rate, caused by the diversity of trigger,
e.g., through movement. DNS has not been designed for such frequent
update, rendering it useless for such highly dynamic applications.
With many edge scenarios in the VR/AR space demanding interactivity
and being latency-sensitive, efficient routing will be key to any
solution.
Various ongoing work on service request forwarding [RFC8677] with the
service function chaining [RFC7665] framework as well as name-based
routing [ICN5G][ICN4G] addresses some aspects described above albeit
with a focus on HTTP as the main invocation protocol. Extensions will
Trossen, et al. Expires April 25, 2021 [Page 13]
INTERNET DRAFT App-Centric Micro-Services
be required to support other invocation protocols, such as GRPC or
MPI (for distributed AI use cases, as outlined in Section 3.3.).
Proposals such as those in [DYN-CAST] suggest extensions to the IP
anycast scheme to enable the flexible routing of service requests to
one or more service instances. Common to those proposals is the use
of a semantic identifier, often a service identifier akin to a URL,
in the routing decision within the network.
5.5 Constraint-based Forwarding Decisions
Allocating the right resources to the right micro-services is a
fundamental task when executing micro-services across highly
distributed micro-data centres (e.g., resource management in cloud
[CLOUDFED]). This is particularly important in the light of volatile
resource availability as well as concurrent and highly dynamic
resource access. Once the specific set of micro-services for an
application has been identified, requirements (e.g., QoS) must be
ensured by the execution environment. Therefore, all micro-data
centres and the execution environment will need to realize mechanisms
for ensuring the utilization of specific resources within a pool of
resources for a specific set of micro-services belonging to one
application, while also ensuring integrity of the wider system.
In relation to the service routing capability discussed in the
previous sub-section, constraints may need to be introduced into the
forwarding decisions for service requests. Such constraints will
likely go beyond network load and latency, as often applied in
scenarios such as load balancing in CDNs. Instead, those constraints
are generally app/service-specific and will need a suitable
representation for the use within network nodes, i.e., the routers
that are forwarding service requests. Moreover, individual router
decisions (e.g., realized through matching operations such as
min/max/equal over a constraint representation) may be coordinated to
achieve a distribution of service requests among many service
instances, effectively realizing a service scheduling capability in
the network, optimized around service-specific constraints, not
unlike many existing data centre service switching schemes.
5.6 Collective Communication
Many micro-service scenarios may exhibit some form of collective
communication beyond 'just' unicast communication, therefore
requiring support for 1:M, M:1, and M:N communication. It is
important to consider here that such collective communication is
often extremely short-lived and can even take place at the level of a
single request, i.e., a following request may exhibit a different
communication pattern, even at least a different receiver group for
the same pattern, such as in the case of an interactive game. It is
Trossen, et al. Expires April 25, 2021 [Page 14]
INTERNET DRAFT App-Centric Micro-Services
therefore required that solutions for supporting such collective
communication must support the spontaneous formation of multicast
relations, as observed in those scenarios.
5.7 State Synchronization
Given the highly distributed nature of app-centric micro-services,
their state exchange and synchronization is a very crucial aspect for
ensuring in-application and system wide consistency. Mechanisms that
ensure consistency will ensure that data is synchronized with
different spatial, temporal and relational data within a given time
period. From the perspective of support through in-network compute
capabilities, such as provided through technologies like P4, it is
important to consider what system and protocol support is required to
utilize such in-network capabilities.
5.8 Dynamic Contracts
NOTE: left for future revision
6 Security Considerations
The use of semantic (or service) identifiers for routing decisions,
as mentioned in Section 5.4October 1, 2018April 4, 2019, requires
methods to ensure the privacy and security of the communication
through avoiding the exposure of service semantic (which is realized
at the application layer) to the network layer, therefore opening up
the opportunity for traffic inspection, among other things. The use
of cryptographic information, e.g., through self-certifying
identifiers, should be investigated to mitigate potential security
and privacy risks.
7 IANA Considerations
N/A
8 Conclusion
This draft positions the evolution of data centres as one of becoming
execution centres for the app-centric experiences provided today
mainly by smart phones directly. With the proliferation of data
centres closer to the end user in the form of edge-based micro data
centres, we believe that app-centric experiences will ultimately be
executed across those many, highly distributed execution points that
this increasingly rich edge environment will provide, such as smart
glasses and IoT devices. Although a number of activities are
currently underway to address some of the challenges for realizing
such AppCentre evolution, we believe that the proposed COIN research
Trossen, et al. Expires April 25, 2021 [Page 15]
INTERNET DRAFT App-Centric Micro-Services
group will provide a suitable forum to drive forward the remaining
research and its dissemination into working systems and the necessary
standardization of key aspects and protocols.
9 References
9.1 Normative References
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, DOI
10.17487/RFC2119, March 1997, <https://www.rfc-
editor.org/info/rfc2119>.
[RFC7665] Halpern, J., Ed., and C. Pignataro, Ed., "Service Function
Chaining (SFC) Architecture", RFC 7665, DOI
10.17487/RFC7665, October 2015, <https://www.rfc-
editor.org/info/rfc7665>.
9.2 Informative References
[MSERVICE1] Dragoni, N., Giallorenzo, S., Lafuente, A. L., Mazzara,
M., Montesi, F., Mustafin, R., & Safina, L. (2017).
Microservices: yesterday,today, and tomorrow. In Present
and Ulterior Software Engineering (pp. 195-216). Springer,
Cham.
[MSERVICE2] Balalaie, A., Heydarnoori, A., & Jamshidi, P. (2016).
Microservices architecture enables devops: Migration to a
cloud-native architecture. IEEE Software, 33(3), 42-52.
[SRVLESS] C. Cicconetti, M. Conti and A. Passarella, "An
Architectural Framework for Serverless Edge Computing:
Design and Emulation Tools," 2018 IEEE International
Conference on Cloud Computing Technology and Science
(CloudCom), Nicosia, 2018, pp. 48-55. doi:
10.1109/CloudCom2018.2018.00024
[TOSCA] Topology and Orchestration Specification for Cloud
Applications Version 1.0. 25 November 2013. OASIS
Standard. <http://docs.oasis-
open.org/tosca/TOSCA/v1.0/os/TOSCA-v1.0-os.html>.
[ERLANG] Armstrong, Joe, et al. "Concurrent programming in ERLANG."
(1993).
[SCOMPOSE] M. Hirzel, R. Soule, S. Schneider, B. Gedik, and R. Grimm,
"A Catalog of Stream Processing Optimizations", ACM
Computing Surveys,46(4):1-34, Mar. 2014
Trossen, et al. Expires April 25, 2021 [Page 16]
INTERNET DRAFT App-Centric Micro-Services
[SDEPLOY1] Lu, H., Shtern, M., Simmons, B., Smit, M., & Litoiu, M.
(2013, June). Pattern-based deployment service for next
generation clouds. In 2013 IEEE Ninth World Congress on
Services (pp. 464-471). IEEE.
[SDEPLOY2] Eilam, T., Elder, M., Konstantinou, A. V., & Snible, E.
(2011, May). Pattern-based composite application
deployment. In 12th IFIP/IEEE International Symposium on
Integrated Network Management (IM 2011) and Workshops (pp.
217-224). IEEE.
[RFC8677] Trossen, D., Purkayastha, D., Rahman, A., "Name-Based
Service Function Forwarder (nSFF) Component within a
Service Function Chaining (SFC) Framework", RFC 8677,
November 2019.
[ICN5G] Ravindran, R., Suthar, P., Trossen, D., Wang, C., White,
G., "Enabling ICN in 3GPP's 5G NextGen Core Architecture",
<https://tools.ietf.org/html/draft-ravi-icnrg-5gc-icn-03>
(work in progress), March 2019.
[ICN4G] Suthar, P., Jangam, Ed., Trossen, D., Ravindran, R.,
"Native Deployment of ICN in LTE, 4G Mobile Networks",
<https://tools.ietf.org/html/draft-irtf-icnrg-icn-lte-4g-
03> (work in progress), March 2019.
[CLOUDFED] M. Liaqat, V. Chang, A. Gani, S. Hafizah Ab Hamid, M.
Toseef, U. Shoaib, R. Liaqat Ali, "Federated cloud
resource management: Review and discussion", Elsevier
Journal of Network and Computer Applications, 2017.
[GRPC] High performance open source universal RPC framework,
https://grpc.io/
[MPI] A. Vishnu, C. Siegel, J. Daily, "Distributed TensorFlow with
MPI", https://arxiv.org/pdf/1603.02339.pdf
[FCDN] M. Al-Naday, M. J. Reed, J. Riihijarvi, D. Trossen, N. Thomos,
M. Al-Khalidi, "fCDN: A Flexible and Efficient CDN
Infrastructure without DNS Redirection of Content
Reflection", https://arxiv.org/pdf/1803.00876.pdf
[DYN-CAST] Y. Li, J. He, L. Geng, P. Liu, Y. Cui, "Framework of
Compute First Networking (CFN)",
<https://tools.ietf.org/html/draft-li-rtgwg-cfn-framework-
00> (work in progress), November 2019
[SA2-5GLAN] 3gpp-5glan, "SP-181129, Work Item Description,
Trossen, et al. Expires April 25, 2021 [Page 17]
INTERNET DRAFT App-Centric Micro-Services
Vertical_LAN(SA2), 5GS Enhanced Support of Vertical and
LAN Services", 3GPP,
<http://www.3gpp.org/ftp/tsg_sa/TSG_SA/Docs/SP-181120.zip>
Authors' Addresses
Dirk Trossen
Huawei Technologies Duesseldorf GmbH
Riesstr. 25C
80992 Munich
Germany
Email: Dirk.Trossen@Huawei.com
Chathura Sarathchandra
InterDigital Europe, Ltd.
64 Great Eastern Street, 1st Floor
London EC2A 3QR
United Kingdom
Email: Chathura.Sarathchandra@InterDigital.com
Michael Boniface
University of Southampton
University Road
Southampton SO17 1BJ
United Kingdom
Email: mjb@it-innovation.soton.ac.uk
Trossen, et al. Expires April 25, 2021 [Page 18]