Use less CPU for slow targets (#36)

This commit re-introduces buffered writing, using less CPU when we have slow
targets. Once we have a need for GB/s range writing we'll need to figure
something else out but for now this is good enough. I'm more interested in
addressing generated data quality.

Signed-off-by: Brian L. Troutwine <brian@troutwine.us>

Author: blt

Cargo.lock

@@ -1,6 +1,6 @@
 [package]
 name = "file_gen"
-version = "0.3.2"
+version = "0.3.3"
 authors = ["Brian L. Troutwine <brian@troutwine.us>"]
 edition = "2018"
 license = "MIT"
@@ -70,6 +70,6 @@ default-features = false
 features = ["rt", "rt-multi-thread", "macros", "fs", "io-util"]
 
 [profile.release]
-# lto = true        # Optimize our binary at link stage.
-# codegen-units = 1 # Increases compile time but improves optmization alternatives.
-# opt-level = 3     # Optimize with 'all' optimization flipped on. May produce larger binaries than 's' or 'z'.
+lto = true        # Optimize our binary at link stage.
+codegen-units = 1 # Increases compile time but improves optmization alternatives.
+opt-level = 3     # Optimize with 'all' optimization flipped on. May produce larger binaries than 's' or 'z'.

Cargo.toml

@@ -54,7 +54,9 @@ throughput as possible. In the 0.3 series we have pivoted to a slower line
 construction mechanism but relied on prebuilding blocks to write, dramatically
 improving total throughput up to the limit of 4Gb/s per target duplicate. This
 is currently a hard limit based on a u32 embedded in the program. See "Weird
-Quirks".
+Quirks". To avoid excessive CPU use for slow targets we currently used buffered
+writing, which appears to limit out at just above 150Mb/s per target, more than
+enough for the needs of the vector project today.
 
 ## Weird Quirks
 
@@ -65,3 +67,6 @@ limitation needs to be lifted we'll have to contribute a fix upstream, or adjust
 our rate limiting approach.
 
 Json generation is painfully slow. I'm very open to alternative approaches.
+
+The use of Arbitrary has been... interesting. You'll find that Ascii generates a
+lot of empty lines.

README.md

@@ -12,7 +12,7 @@ use std::convert::TryInto;
 use std::num::{NonZeroU32, NonZeroU64};
 use std::path::PathBuf;
 use tokio::fs;
-use tokio::io::AsyncWriteExt;
+use tokio::io::{AsyncWriteExt, BufWriter};
 
 #[derive(Debug)]
 pub enum Error {
@@ -200,18 +200,21 @@ impl Log {
             &labels
         );
 
-        let mut fp = fs::OpenOptions::new()
-            .create(true)
-            .truncate(false)
-            .write(true)
-            .open(&self.path)
-            .await?;
+        let mut fp = BufWriter::with_capacity(
+            ONE_MEBIBYTE * 100,
+            fs::OpenOptions::new()
+                .create(true)
+                .truncate(false)
+                .write(true)
+                .open(&self.path)
+                .await?,
+        );
 
         for (total_bytes, total_newlines, block) in self.block_cache.iter().cycle() {
             self.rate_limiter.until_n_ready(*total_bytes).await?;
 
             {
-                fp.write(block).await?;
+                fp.write_all(block).await?;
                 // block.len() and total_bytes are the same numeric value but we
                 // avoid needing to get a plain value from a non-zero by calling
                 // len here.
@@ -231,12 +234,15 @@ impl Log {
                 // Open a new fp to `self.path`, replacing `fp`. Any holders of
                 // the file pointer still have it but the file no longer has a
                 // name.
-                fp = fs::OpenOptions::new()
-                    .create(true)
-                    .truncate(false)
-                    .write(true)
-                    .open(&self.path)
-                    .await?;
+                fp = BufWriter::with_capacity(
+                    ONE_MEBIBYTE * 100,
+                    fs::OpenOptions::new()
+                        .create(true)
+                        .truncate(false)
+                        .write(true)
+                        .open(&self.path)
+                        .await?,
+                );
                 bytes_written = 0;
                 counter!("file_rotated", 1, &labels);
             }

src/file.rs

@@ -23,7 +23,7 @@ impl<'a> Arbitrary<'a> for Member {
     }
 
     fn size_hint(_depth: usize) -> (usize, Option<usize>) {
-        (100, Some(6144)) // 100B to 6KiB
+        (0, Some(6144)) // 100B to 6KiB
     }
 }