... | ... | @@ -278,7 +278,7 @@ The enhancements to current IEEE 802.1AS include |
|
|
- SPB+SRP: use Shortest Path Bridging to identify
|
|
|
different paths, then Stream Reservation Protocol
|
|
|
|
|
|
## **AVB *Gen2* vs. White Rabbit**
|
|
|
## **White Rabbit**
|
|
|
|
|
|
### Requrements
|
|
|
|
... | ... | @@ -294,10 +294,10 @@ The enhancements to current IEEE 802.1AS include |
|
|
500-5000 bytes (encoded into several frames of size 300-1500
|
|
|
bytes and sent in burst,
|
|
|
[FEC](https://www.ohwr.org/project/white-rabbit/uploads/ff98688542afbe36e9f11877ba4c5f1a/ICALEPCS2011_poster.pdf)),
|
|
|
~8 "control streams" (defined in VLANs, see
|
|
|
~8 "control streams" (defined within separate VLANs, see
|
|
|
[page 7](https://www.ohwr.org/project/white-rabbit/uploads/1244844e3c3cf850fad84e575f3b3f61/wrCernControlAndTiming.v1.1.pdf))
|
|
|
sent every 1000us, one-to-many, "normal data" size
|
|
|
(payload): ~1500 bytes
|
|
|
sent every 1000us, critical stream is one-to-many, "normal
|
|
|
data" size (payload): ~1500 bytes
|
|
|
- GSI (not necessarily up-to-date)
|
|
|
- maximum latency: 100us over ~4 bridge hops @ 1Gbps
|
|
|
- guaranteed latency over tree-like topology
|
... | ... | @@ -307,9 +307,11 @@ The enhancements to current IEEE 802.1AS include |
|
|
200-500 bytes (encoded into 4 frames of size 150-300 bytes
|
|
|
and sent in burst,
|
|
|
[FEC](https://www.ohwr.org/project/white-rabbit/uploads/ff98688542afbe36e9f11877ba4c5f1a/ICALEPCS2011_poster.pdf)),
|
|
|
sent every 100us, one-to-many, "normal data" size (payload):
|
|
|
~1500 bytes
|
|
|
- Synchronization: sub-nanosecond in the entire network
|
|
|
sent every 100us, critical stream is one-to-many, "normal
|
|
|
data" size (payload): ~1500 bytes
|
|
|
- Synchronization:
|
|
|
- accuracy: sub-nanosecond in the entire network
|
|
|
- precision: order of picoseconds in the entire network
|
|
|
- Reliability seamless redundancy (synchronization-wise and critical
|
|
|
data-wise)
|
|
|
|
... | ... | @@ -317,20 +319,12 @@ The enhancements to current IEEE 802.1AS include |
|
|
|
|
|
- profile/extension of PTP: WRPTP
|
|
|
- requires physical syntonization using Synchronous Ethernet
|
|
|
- using phase detection (DDMTD) to enhance HW-timestamping precision
|
|
|
to picosecond level
|
|
|
- uses phase detection (DDMTD) to enhance HW-timestamping precision to
|
|
|
picosecond level
|
|
|
- auto-calibration of link asymmetry (single fiber and DDMTD phase
|
|
|
detection)
|
|
|
- support of network redundancy and seamless (no time) reconfiguration
|
|
|
|
|
|
*AVBg2 vs. WR:**
|
|
|
The main difference is the usage of SyncE and single fiber in WR. AVBg2
|
|
|
bridges will have to be time aware, so will probably syntonize to
|
|
|
grandmaster's frequency using Sync Messages and will need to synchronize
|
|
|
to grandmaster, like WR (AVBg1 bridges are transparent and syntonize
|
|
|
optionally). The efforts to support network redundancy and speed-up
|
|
|
reconfiguration seem to be inline in WR and AVB. (**ToFinish**)
|
|
|
|
|
|
### Reliability
|
|
|
|
|
|
- optional network redundancy
|
... | ... | @@ -379,3 +373,27 @@ reconfiguration seem to be inline in WR and AVB. (**ToFinish**) |
|
|
- some problem when cut-through for non-critical traffic -- we
|
|
|
don't know the size of the frame
|
|
|
|
|
|
## AVBg2 vs. WR
|
|
|
|
|
|
### Synchronization-wise
|
|
|
|
|
|
The main difference is the usage of SyncE and single fiber in WR. AVBg2
|
|
|
bridges will have to be time aware, so will probably syntonize to
|
|
|
grandmaster's frequency using Sync Messages and will need to synchronize
|
|
|
to grandmaster, like WR (AVBg1 bridges are transparent and syntonize
|
|
|
optionally). The efforts to support network redundancy and speed-up
|
|
|
reconfiguration seem to be inline in WR and AVB. (**ToFinish**)
|
|
|
|
|
|
### Critical-data-wise
|
|
|
|
|
|
WR does not foresee dynamic stream reservation -- it is assumed that
|
|
|
there are two kinds of (1) critical data (so called Control Messages,
|
|
|
CMs) and (2) non-critical data (Best-effort, BE). All critical data is
|
|
|
treated equally and share the same resources which are allocated to it.
|
|
|
The characteristics distribution of the CMs in WR is also different then
|
|
|
in the distribution of streams in AVB. In WR the critical data stream is
|
|
|
supposed to be broadcast within VLAN, thus a stream is propagated to all
|
|
|
the end stations within a VLAN: a particular case of such solution is to
|
|
|
define VLAN for a stream between two end stations. Still, separate
|
|
|
VLAN-streams do not have separate resources.
|
|
|
|