View Revisions: Issue #64
| Summary | 0000064: implement swarm | ||
|---|---|---|---|
| Revision | 2022-05-15 09:03 by ralph | ||
| Description | Hard requirements: * participant -> swarm request <- answer * swarm -> participant request <- answer * swarm can split if physically disrupted * swarm can merge if physically connected Soft requirements: * swarm participants share a relative common base time Option 1: * bases on multiple channels * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are an abstract rating, logical channel distance (hash distance for opendht) as well as latency (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 09:03 by ralph | ||
| Description | Hard requirements: * participant -> swarm request <- answer * swarm -> participant request <- answer * swarm can split if physically disrupted Soft requirements: * swarm participants share a relative common base time Option 1: * bases on multiple channels * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are an abstract rating, logical channel distance (hash distance for opendht) as well as latency (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 09:01 by ralph | ||
| Description | * participant -> swarm request <- answer * swarm -> participant request <- answer * swarm can split if physically disrupted * swarm participants share a relative common base time Option 1: * bases on multiple channels * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are an abstract rating, logical channel distance (hash distance for opendht) as well as latency (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 09:01 by ralph | ||
| Description | * content is immutable (first version) * participant -> swarm request <- answer * swarm -> participant request <- answer * swarm can split if physically disrupted * swarm participants share a relative common base time Option 1: * bases on multiple channels * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are an abstract rating, logical channel distance (hash distance for opendht) as well as latency (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 09:01 by ralph | ||
| Description | * content is immutable (first version) * bases on multiple channels * participant -> swarm request <- answer * swarm -> participant request <- answer * swarm can split if physically disrupted * swarm participants share a relative common base time Option 1: * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are an abstract rating, logical channel distance (hash distance for opendht) as well as latency (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 09:00 by ralph | ||
| Description | * content is immutable (first version) * bases on multiple channels * participant -> swarm request <- answer * swarm -> participant request <- answer * swarm can split if physically disrupted * swarm share a relative common base time Option 1: * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are an abstract rating, logical channel distance (hash distance for opendht) as well as latency (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 08:59 by ralph | ||
| Description | * content is immutable (first version) * bases on multiple channels * participant -> swarm request <- answer * swarm -> participant request <- answer * swarm can split if physically disrupted Option 1: * masters share a relative common base time * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are an abstract rating, logical channel distance (hash distance for opendht) as well as latency (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 08:57 by ralph | ||
| Description | * content is immutable (first version) * bases on multiple channels * participant -> swarm request <- answer * swarm -> participant request <- answer * masters share a relative common base time * masters are executing swam integrity poll in given intervals * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are an abstract rating, logical channel distance (hash distance for opendht) as well as latency (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests and chooses a new master * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 08:56 by ralph | ||
| Description | * content is immutable (first version) * bases on multiple channels * participant -> swarm request <- answer * swarm -> participant request <- answer * masters share a relative common base time * masters are executing swam integrity poll in given intervals * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are logical channel distance (hash distance for opendht) as well as latency and an abstract rating (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests and chooses a new master * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 08:55 by ralph | ||
| Description | * content is immutable (first version) * bases on multiple channels * participant -> swarm request <- answer * swarm -> participant request <- answer * masters share a relative common base time * masters are executing swam integrity poll in given intervals * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are logical channel distance (hash distance for opendht) as well as latency and content completeness (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests and chooses a new master * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 08:55 by ralph | ||
| Description | * content is immutable (first version) * bases on multiple channels * participant -> swarm information < ack * participant -> swarm request <- answer * swarm -> participant information <- ack * swarm -> participant request <- answer * masters share a relative common base time * masters are executing swam integrity poll in given intervals * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are logical channel distance (hash distance for opendht) as well as latency and content completeness (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests and chooses a new master * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-15 08:53 by ralph | ||
| Description | * content is immutable (first version) * bases on multiple channels * participant -> swarm information * participant -> swarm request <- answer * swarm -> participant information * swarm -> participant request <- answer * masters share a relative common base time * masters are executing swam integrity poll in given intervals * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are logical channel distance (hash distance for opendht) as well as latency and content completeness (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests and chooses a new master * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-13 13:15 by ralph | ||
| Description | * content is immutable (first version) * bases on multiple channels * masters share a relative common base time * masters are executing swam integrity poll in given intervals * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are logical channel distance (hash distance for opendht) as well as latency and content completeness (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests and chooses a new master * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-13 13:10 by ralph | ||
| Description | * messages are immutable (first version) * bases on multiple channels * masters share a relative common base time * masters are executing swam integrity poll in given intervals * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are logical channel distance (hash distance for opendht) as well as latency and content completeness (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests and chooses a new master * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-13 13:07 by ralph | ||
| Description | * bases on multiple channels * masters share a relative common base time * masters are executing swam integrity poll in given intervals * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are logical channel distance (hash distance for opendht) as well as latency and content completeness (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests and chooses a new master * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
| Revision | 2022-05-13 13:05 by ralph | ||
| Description | * masters share a relative common base time * masters are executing swam integrity poll in given intervals * a master redundancy ratio is to be maintained * if required there is a tendering for master roles * qualifications are logical channel distance (hash distance for opendht) as well as latency and content completeness (the furthest from all other masters who made a bid most quickly) * a slave chooses nearest master to serve it's requests * a slave blames a master to the other masters if it's not answering requests and chooses a new master * if new master if not answering too, it chooses the next one until there are no masters left to try, then assumes it's disconnected from swarm * on swarm split visibility requests are sent and relay roles assigned * on swam merge obsolete relay roles are dissolved |
||
