From 6afa79a5a501787c2db4f25ae337cc0bc68e967c Mon Sep 17 00:00:00 2001 From: kaichaosun Date: Wed, 25 Sep 2024 19:19:28 +0800 Subject: [PATCH] options for mvds --- .../application/bandwidth-optimization.md | 24 +++++++++++++++++-- 1 file changed, 22 insertions(+), 2 deletions(-) diff --git a/standards/application/bandwidth-optimization.md b/standards/application/bandwidth-optimization.md index 093e4df..1a315d5 100644 --- a/standards/application/bandwidth-optimization.md +++ b/standards/application/bandwidth-optimization.md @@ -60,9 +60,29 @@ _Concerns:_ ### Message Retransmission in MVDS -Messages are resent when the recipient is offline, leading to additional network load. +Direct messages, group messages and some community join request messages are sent via MVDS. When the recipient is offline, these messages are repeatedly resent until a predefined limit is reached, which leads to more bandwidth usage. + +**Option 1** + +Replace MVDS with a more reliable and bandwidth-efficient end-to-end (E2E) [reliability protocol](https://forum.vac.dev/t/end-to-end-reliability-for-scalable-distributed-logs/293) to reduce message retransmissions. + +_Concerns:_ +- fully adopting e2e reliability protocol requires significant time and resources for implementation. + +**Option 2** + +Disable MVDS retransmission for recipients that are not online/active. + +_Concerns:_ +- user may set their status to offline but still want to receive messages. (TODO needs input from Status team) + +**Option 3** + +Increase the time interval between message retransmissions to reduce bandwidth usage. + +_Concerns:_ +- it may increase the latency of message delivery. -Replace MVDS with a more reliable and bandwidth-efficient end-to-end (E2E) reliability protocol to reduce redundant message retransmissions. ### Updates for Descriptive Messages