[Css-csts] Service Management of SLE frame buffer sizes
John Pietras
john.pietras at gst.com
Thu Oct 30 09:13:32 EST 2008
CSTSWG and SMWG colleagues ---
During the Berlin meeting Wolfgang Hell and I met to ensure that the
service management aspects of SLE transfer services were consistent
across the SLE Transfer Service Specifications and the SCCS-SM Service
Specification. We discovered a few items that required adjustment in
either the SLE TS or SCCS-SM red books, and have begun making those
adjustments.
I have just come across a few more items regarding frame buffer sizes
that did not surface in our meeting. I hope that we can come to a
consensus as to how to handle these items in the SLE TS and/or SCCS-SM
books.
Sections 2.6.4.6.3 (Online Frame Buffer and Offline Frame Buffer) of the
RAF and RCF specifications state "In the case of complete online
delivery mode, the online frame buffer is intended to overcome
limitations of the communications service: bandwidth limitations,
outages, and congestion. In the case of offline delivery mode, the
offline frame buffer is intended to enable data to be delivered hours or
days after the completion of the space link session. The exact size of
these buffers is set by service management. It is normally expected
that the online frame buffer is sufficiently large to hold all data that
might be accumulated during one space link session and that the offline
frame buffer is sufficiently large to hold all data that might be
accumulated during several space link sessions."
Although this paragraph states that Service Management is responsible
for setting these buffer sizes, there are no corresponding entries in
the corresponding tables 3-1 (Setting of XXX Service Configuration
Parameters). I believe that they should be added as items under the
Service Management column.
Regarding how they should be addressed by SM, at what level of
granularity should they be managed?
A. Online Frame Buffers
Can one online frame buffer size be specified in the Service Agreement
and used for all online transfer service instances, or does each
transfer service instance need to be able to specify its own buffer
size? If each transfer service instance needs to be able to specify its
own online buffer size, should there be a maximum size in the Service
Agreement that constrains all individual service instance online buffer
sizes? (Currently, the SCCS-SM Red Book has a single global size
specified in the Service Agreement).
B. Offline Frame Buffers
Because offline buffers are (by definition) persistent beyond the
boundaries of the Service Package in which the data were received,
offline buffers cannot be directly equated to individual offline SLE
transfer service instances (nor to complete CSTS service instances, when
those become available). Instead, SCCS-SM employs the concept of
persistent data stores associated with the ground stations (technically,
with the antennas). Each Complex antenna (e.g., a ground station
antenna) identified in the Service Agreement as being able to support
the mission has associated with it a data store that can hold the data
collected via that antenna for that mission. Space Link Session Service
Packages (i.e., Service Packages that involve active space links) may
include data sink instances that filter the data to be stored from the
space link session (e.g., all frames from symbol stream X, VCs 2 and 4
from symbol stream Y). Each data sink instances also associates the
stored data with a functional group ID that qualifies the data. [Note
that this is a more dynamic use of the functional group ID than is
currently made in the SLE Transfer Services, and is more in line with
early 1990s concepts of how the functional group ID might be used]
Retrieval Service Packages establish offline transfer service instances
that are associated with a specific antenna and that can retrieve any
data stored therein. The functional group ID that is specified as part
of the service instance identifier of the offline transfer service
instance also qualifies what data can be retrieved by the transfer
service instance, i.e., the transfer service instance can only retrieve
the stored data that was "tagged" (by the data sink) with the functional
group ID that matches the functional group ID of that transfer service
instance.
The SCCS-SM Red-3 Book has the features to support this approach for
storage and retrieval of offline data, but it does not yet specify the
offline buffer sizes, an omission that must be corrected for Blue-1.The
approach that I propose to remedy this omission is to associate a
maximum data store capacity (in GigaOctets) with each allowed antenna
specified in the Service Agreement. Is this approach acceptable?
I am looking forward to your comments.
Best regards,
John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ccsds.org/pipermail/css-csts/attachments/20081030/ff263f3a/attachment.htm
More information about the Css-csts
mailing list