Benchmark skeletondoc/benchmarks

8 files changed, 105 insertions, 0 deletions

diff --git a/doc/book/benchmarks/_index.md b/doc/book/benchmarks/_index.md
new file mode 100644
index 00000000..f4e8edb6
--- /dev/null
+++ b/doc/book/benchmarks/_index.md
@@ -0,0 +1,8 @@
++++
+title = "Benchmarks"
+weight = 60
+sort_by = "weight"
+template = "documentation.html"
++++
+
+Hello
diff --git a/doc/book/benchmarks/abstraction.md b/doc/book/benchmarks/abstraction.md
new file mode 100644
index 00000000..a429c787
--- /dev/null
+++ b/doc/book/benchmarks/abstraction.md
@@ -0,0 +1,17 @@
++++
+title = "Abstraction cost"
+weight = 30
++++
+
+We take as our baseline the raw disk sequential write performance.
+We then compare Garage's performances to it, the difference represents what we call our "abstraction cost".
+
+fsync, chunking, compression, pipelining, synchronization
+
+# raw perf VS garage
+
+# garage tmpfs VS garage std
+
+# garage various multipart sizes
+
+# garage 0.7.2.1 VS upstream
diff --git a/doc/book/benchmarks/failure_recovery.md b/doc/book/benchmarks/failure_recovery.md
new file mode 100644
index 00000000..59c9399d
--- /dev/null
+++ b/doc/book/benchmarks/failure_recovery.md
@@ -0,0 +1,18 @@
++++
+title = "Failure & recovery"
+weight = 50
++++
+
+# Failure impact
+
+Failures will lead to timeouts, which in turn could
+lead to failed requests (this is a bug if failure enters in Garage tolerance)
+and to increased latency as some retries might be performed.
+
+How we proceed: we pause (`kill -STOP xxx`) one Garage process.
+The idea is we don't want to close the TCP connection that would signal too easily
+that a crash occured. Instead, we want to simulate a network error
+or an overloaded process, ie. a 'non-collaborating' crash.
+
+
+# Recovery impact
diff --git a/doc/book/benchmarks/industry_tools.md b/doc/book/benchmarks/industry_tools.md
new file mode 100644
index 00000000..99b832e4
--- /dev/null
+++ b/doc/book/benchmarks/industry_tools.md
@@ -0,0 +1,14 @@
++++
+title = "Industry tools"
+weight = 60
++++
+
+
+# minio warp 
+
+# intel-cloud cosbench
+
+# (non retenu)
+
+ - wasabi s3-benchmark
+ - https://github.com/dvassallo/s3-benchmark
diff --git a/doc/book/benchmarks/liveness.md b/doc/book/benchmarks/liveness.md
new file mode 100644
index 00000000..35ea19b5
--- /dev/null
+++ b/doc/book/benchmarks/liveness.md
@@ -0,0 +1,10 @@
++++
+title = "Liveness"
+weight = 40
++++
+
+freedom from starvation, backpressure, etc.
+
+# Responsiveness under read/write load
+
+
diff --git a/doc/book/benchmarks/network_sensitiveness.md b/doc/book/benchmarks/network_sensitiveness.md
new file mode 100644
index 00000000..2679811f
--- /dev/null
+++ b/doc/book/benchmarks/network_sensitiveness.md
@@ -0,0 +1,14 @@
++++
+title = "Network sensitiveness"
+weight = 10
++++
+
+impact of node count and their latency
+
+# Latency amplification
+
+# Cluster size impact
+
+# Time-To-First-Byte (TTFB)
+
+with various object size
diff --git a/doc/book/benchmarks/pushing_limits.md b/doc/book/benchmarks/pushing_limits.md
new file mode 100644
index 00000000..fc0f0a72
--- /dev/null
+++ b/doc/book/benchmarks/pushing_limits.md
@@ -0,0 +1,16 @@
++++
+title = "Pushing limits"
+weight = 60
++++
+
+# Many objects
+
+# Huge objects
+
+# Many nodes (horizontal scalability)
+
+
+
+# Large nodes (vertical scalability)
+
+
diff --git a/doc/book/benchmarks/real_world.md b/doc/book/benchmarks/real_world.md
new file mode 100644
index 00000000..741e6c7f
--- /dev/null
+++ b/doc/book/benchmarks/real_world.md
@@ -0,0 +1,8 @@
++++
+title = "Real world"
+weight = 80
++++
+
+# Nextcloud
+
+# Peertube