blob: e1c65d2f5cc828c265ded386bc72a07e0d198094 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
|
In doc and code, convert JSON-RPC param "parameters" to "arguments" in order to avoid
confusion with JSON-RPC "params"
In doc and code, rename "target" "service" or potentially "target
service"
In HLD, update data_link_up use case (Step 10) to involve service
discovery
In HLD, schedule, data link, and service discovery - Fix code so that
service availability is managed by service discovery only. Today
the network address of a given service (set by data_link_up calls
from data_link_device) are received by schedule, which maintains a
redundant copy of all service availability states. This should be
moved to Servide Discovery, which should forward the call to
schedule. HLD
All repos
Update README.md.
Remove find_setup.sh
Fix config.
Prune xxx_lib.erl
Service Discovery
Node service prefix matching should return the one with the longest
prefix match. This allows for default static nodes to be setup with
only a root prefix that can be used as a fallback.
Service Discovery, Service Edge, Data Link
If a new local service connects and registers with a Service Edge,
the service should be announced to other RVI nodes that are
currently connected to the local node. (Real time service
announcement)
Scheduler
If two different services are invoked on a remote node that is
currently not available, the services may be invoked out of order
when the node does become connected. Scheduluer should queue
messages on a per-node basis, not per-service.
FIXED:
Data Link
Any broken connection to a statically configured node should
periodically get a reconnect attempt.
|
