From 0793713f1ce118fe4741a5e8f31b93445909e188 Mon Sep 17 00:00:00 2001 From: Propfend Date: Sat, 27 Dec 2025 19:12:44 -0300 Subject: [PATCH 1/3] add paddler faq --- README.md | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/README.md b/README.md index ab4a32d7..549499de 100644 --- a/README.md +++ b/README.md @@ -102,6 +102,17 @@ And use a GUI to test the inference: * [Create a multi agent fleet](https://paddler.intentee.com/docs/starting-out/multi-agent-fleet/) * [Go beyond a single device](https://paddler.intentee.com/docs/starting-out/going-beyond-a-single-device/) +## FAQ +* What’s the difference between Paddler and llama.cpp RPC? + - Paddler has: + - Dynamic service discovery, allowing agents to be added or removed at runtime + - Fault tolerance, with automatic handling of unavailable agents + - LLM-aware load balancing, based on slot utilization + - Runtime reconfiguration, without requiring restarts + - Elastic scaling, including scaling to and from zero hosts through request buffering + - In contrast, llama.cpp RPC is focused on practicity, relies on predefined hosts configured on start-up, offers limited fault tolerance, and uses round-robin load-balancing strategy that is unaware of LLM execution time. + For more information, see [GitHub issue #35](https://github.com/intentee/paddler/issues/35). + ## Why the Name We initially wanted to use [Raft](https://raft.github.io/) consensus algorithm (thus Paddler, because it paddles on a Raft), but eventually dropped that idea. The name stayed, though. From c26fd03b6460291c328dade0fb0b161ce822ca34 Mon Sep 17 00:00:00 2001 From: Luiz Miguel <120745901+Propfend@users.noreply.github.com> Date: Sat, 27 Dec 2025 19:15:10 -0300 Subject: [PATCH 2/3] Update README.md Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Signed-off-by: Luiz Miguel <120745901+Propfend@users.noreply.github.com> --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 549499de..85fe3d23 100644 --- a/README.md +++ b/README.md @@ -110,7 +110,7 @@ And use a GUI to test the inference: - LLM-aware load balancing, based on slot utilization - Runtime reconfiguration, without requiring restarts - Elastic scaling, including scaling to and from zero hosts through request buffering - - In contrast, llama.cpp RPC is focused on practicity, relies on predefined hosts configured on start-up, offers limited fault tolerance, and uses round-robin load-balancing strategy that is unaware of LLM execution time. + - In contrast, llama.cpp RPC is focused on practicality, relies on predefined hosts configured on start-up, offers limited fault tolerance, and uses round-robin load-balancing strategy that is unaware of LLM execution time. For more information, see [GitHub issue #35](https://github.com/intentee/paddler/issues/35). ## Why the Name From b0a97a62b8683d3d8709d2533a7b958b64ee52b8 Mon Sep 17 00:00:00 2001 From: Propfend Date: Tue, 6 Jan 2026 10:07:01 -0300 Subject: [PATCH 3/3] qustion in expandable, change wording, gh linnk is a list item. --- README.md | 26 +++++++++++++++++--------- 1 file changed, 17 insertions(+), 9 deletions(-) diff --git a/README.md b/README.md index 85fe3d23..191928ed 100644 --- a/README.md +++ b/README.md @@ -103,15 +103,23 @@ And use a GUI to test the inference: * [Go beyond a single device](https://paddler.intentee.com/docs/starting-out/going-beyond-a-single-device/) ## FAQ -* What’s the difference between Paddler and llama.cpp RPC? - - Paddler has: - - Dynamic service discovery, allowing agents to be added or removed at runtime - - Fault tolerance, with automatic handling of unavailable agents - - LLM-aware load balancing, based on slot utilization - - Runtime reconfiguration, without requiring restarts - - Elastic scaling, including scaling to and from zero hosts through request buffering - - In contrast, llama.cpp RPC is focused on practicality, relies on predefined hosts configured on start-up, offers limited fault tolerance, and uses round-robin load-balancing strategy that is unaware of LLM execution time. - For more information, see [GitHub issue #35](https://github.com/intentee/paddler/issues/35). +
+ What’s the difference between Paddler and llama.cpp RPC? + +- **Paddler provides:** + - Dynamic service discovery (agents can be added/removed at runtime) + - Fault tolerance with automatic handling of unavailable agents + - LLM-aware load balancing based on slot utilization + - Runtime reconfiguration without restarts + - Elastic scaling, including scale-to-zero via request buffering + +- **llama.cpp RPC** focuses on practicality, relies on predefined hosts at startup, + has limited fault tolerance, and uses a round-robin load balancing unaware of LLM execution time. + +- For more information, see + [GitHub issue #35](https://github.com/intentee/paddler/issues/35). + +
## Why the Name