public inbox for pbs-devel@lists.proxmox.com
 help / color / mirror / Atom feed
From: Christian Ebner <c.ebner@proxmox.com>
To: pbs-devel@lists.proxmox.com
Subject: [pbs-devel] [RFC v2 proxmox-backup 34/36] fix #3174: client: pxar: enable caching and meta comparison
Date: Tue,  5 Mar 2024 10:27:01 +0100	[thread overview]
Message-ID: <20240305092703.126906-35-c.ebner@proxmox.com> (raw)
In-Reply-To: <20240305092703.126906-1-c.ebner@proxmox.com>

Add the final glue logic to enable the look-ahead caching and
metadata comparison introduced in the preparatory patches.

Signed-off-by: Christian Ebner <c.ebner@proxmox.com>
---
changes since version 1:
- fix pxar exclude cli entry caching

 pbs-client/src/pxar/create.rs | 121 +++++++++++++++++++++++++++++++---
 1 file changed, 113 insertions(+), 8 deletions(-)

diff --git a/pbs-client/src/pxar/create.rs b/pbs-client/src/pxar/create.rs
index b2ce898f..bb4597bc 100644
--- a/pbs-client/src/pxar/create.rs
+++ b/pbs-client/src/pxar/create.rs
@@ -32,10 +32,14 @@ use pbs_datastore::dynamic_index::{
 };
 
 use crate::inject_reused_chunks::InjectChunks;
+use crate::pxar::look_ahead_cache::{CacheEntry, CacheEntryData};
 use crate::pxar::metadata::errno_is_unsupported;
 use crate::pxar::tools::assert_single_path_component;
 use crate::pxar::Flags;
 
+const MAX_CACHE_SIZE: usize = 512;
+const CACHED_PAYLOAD_THRESHOLD: u64 = 2 * 1024 * 1024;
+
 #[derive(Default)]
 struct ReusedChunks {
     start_boundary: PayloadOffset,
@@ -253,6 +257,9 @@ struct Archiver {
     reused_chunks: ReusedChunks,
     previous_payload_index: Option<DynamicIndexReader>,
     forced_boundaries: Arc<Mutex<VecDeque<InjectChunks>>>,
+    cached_entries: Vec<CacheEntry>,
+    caching_enabled: bool,
+    cached_payload_size: u64,
 }
 
 type Encoder<'a, T> = pxar::encoder::aio::Encoder<'a, T>;
@@ -335,16 +342,32 @@ where
         reused_chunks: ReusedChunks::new(),
         previous_payload_index,
         forced_boundaries,
+        cached_entries: Vec::new(),
+        caching_enabled: false,
+        cached_payload_size: 0,
     };
 
     archiver
         .archive_dir_contents(&mut encoder, accessor, source_dir, true)
         .await?;
+
+    if let Some(last) = archiver.cached_entries.pop() {
+        match last {
+            // do not close final directory, this is done by the caller
+            CacheEntry::DirEnd => {}
+            _ => archiver.cached_entries.push(last),
+        }
+    }
+
+    archiver
+        .flush_cached_to_archive(&mut encoder, true, false)
+        .await?;
+
     encoder.finish().await?;
     Ok(())
 }
 
-struct FileListEntry {
+pub(crate) struct FileListEntry {
     name: CString,
     path: PathBuf,
     stat: FileStat,
@@ -396,8 +419,15 @@ impl Archiver {
                 let file_name = file_entry.name.to_bytes();
 
                 if is_root && file_name == b".pxarexclude-cli" {
-                    self.encode_pxarexclude_cli(encoder, &file_entry.name, old_patterns_count)
-                        .await?;
+                    if self.caching_enabled {
+                        self.cached_entries.push(CacheEntry::PxarExcludeCliEntry(
+                            file_entry,
+                            old_patterns_count,
+                        ));
+                    } else {
+                        self.encode_pxarexclude_cli(encoder, &file_entry.name, old_patterns_count)
+                            .await?;
+                    }
                     continue;
                 }
 
@@ -413,6 +443,11 @@ impl Archiver {
                 .await
                 .map_err(|err| self.wrap_err(err))?;
             }
+
+            if self.caching_enabled {
+                self.cached_entries.push(CacheEntry::DirEnd);
+            }
+
             self.path = old_path;
             self.entry_counter = entry_counter;
             self.patterns.truncate(old_patterns_count);
@@ -693,8 +728,6 @@ impl Archiver {
         c_file_name: &CStr,
         stat: &FileStat,
     ) -> Result<(), Error> {
-        use pxar::format::mode;
-
         let file_mode = stat.st_mode & libc::S_IFMT;
         let open_mode = if file_mode == libc::S_IFREG || file_mode == libc::S_IFDIR {
             OFlag::empty()
@@ -732,6 +765,71 @@ impl Archiver {
             self.skip_e2big_xattr,
         )?;
 
+        if self.previous_payload_index.is_none() {
+            return self
+                .add_entry_to_archive(encoder, accessor, c_file_name, stat, fd, &metadata)
+                .await;
+        }
+
+        // Avoid having to many open file handles in cached entries
+        if self.cached_entries.len() > MAX_CACHE_SIZE {
+            self.flush_cached_to_archive(encoder, false, true).await?;
+        }
+
+        if metadata.is_regular_file() {
+            self.cache_or_flush_entries(encoder, accessor, c_file_name, stat, fd, &metadata)
+                .await
+        } else {
+            if self.caching_enabled {
+                if stat.st_mode & libc::S_IFMT == libc::S_IFDIR {
+                    let fd_clone = fd.try_clone()?;
+                    let cache_entry = CacheEntry::DirEntry(CacheEntryData::new(
+                        fd,
+                        c_file_name.into(),
+                        stat.clone(),
+                        metadata.clone(),
+                        PayloadOffset::default(),
+                    ));
+                    self.cached_entries.push(cache_entry);
+
+                    let dir = Dir::from_fd(fd_clone.into_raw_fd())?;
+                    self.add_directory(encoder, accessor, dir, c_file_name, &metadata, stat)
+                        .await?;
+
+                    if let Some(ref catalog) = self.catalog {
+                        if !self.caching_enabled {
+                            catalog.lock().unwrap().end_directory()?;
+                        }
+                    }
+                } else {
+                    let cache_entry = CacheEntry::RegEntry(CacheEntryData::new(
+                        fd,
+                        c_file_name.into(),
+                        stat.clone(),
+                        metadata,
+                        PayloadOffset::default(),
+                    ));
+                    self.cached_entries.push(cache_entry);
+                }
+                Ok(())
+            } else {
+                self.add_entry_to_archive(encoder, accessor, c_file_name, stat, fd, &metadata)
+                    .await
+            }
+        }
+    }
+
+    async fn add_entry_to_archive<T: SeqWrite + Send>(
+        &mut self,
+        encoder: &mut Encoder<'_, T>,
+        accessor: &mut Option<Directory<LocalDynamicReadAt<RemoteChunkReader>>>,
+        c_file_name: &CStr,
+        stat: &FileStat,
+        fd: OwnedFd,
+        metadata: &Metadata,
+    ) -> Result<(), Error> {
+        use pxar::format::mode;
+
         let file_name: &Path = OsStr::from_bytes(c_file_name.to_bytes()).as_ref();
         match metadata.file_type() {
             mode::IFREG => {
@@ -781,7 +879,9 @@ impl Archiver {
                     .add_directory(encoder, accessor, dir, c_file_name, &metadata, stat)
                     .await;
                 if let Some(ref catalog) = self.catalog {
-                    catalog.lock().unwrap().end_directory()?;
+                    if !self.caching_enabled {
+                        catalog.lock().unwrap().end_directory()?;
+                    }
                 }
                 result
             }
@@ -1132,7 +1232,9 @@ impl Archiver {
     ) -> Result<(), Error> {
         let dir_name = OsStr::from_bytes(dir_name.to_bytes());
 
-        encoder.create_directory(dir_name, metadata).await?;
+        if !self.caching_enabled {
+            encoder.create_directory(dir_name, metadata).await?;
+        }
 
         let old_fs_magic = self.fs_magic;
         let old_fs_feature_flags = self.fs_feature_flags;
@@ -1172,7 +1274,10 @@ impl Archiver {
         self.fs_feature_flags = old_fs_feature_flags;
         self.current_st_dev = old_st_dev;
 
-        encoder.finish().await?;
+        if !self.caching_enabled {
+            encoder.finish().await?;
+        }
+
         result
     }
 
-- 
2.39.2





  parent reply	other threads:[~2024-03-05  9:28 UTC|newest]

Thread overview: 94+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-05  9:26 [pbs-devel] [RFC pxar proxmox-backup 00/36] fix #3174: improve file-level backup Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 01/36] format/examples: add PXAR_PAYLOAD_REF entry header Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 02/36] encoder: add optional output writer for file payloads Christian Ebner
2024-03-11 13:21   ` Fabian Grünbichler
2024-03-11 13:50     ` Christian Ebner
2024-03-11 15:41       ` Fabian Grünbichler
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 03/36] format/decoder: add method to read payload references Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 04/36] decoder: add optional payload input stream Christian Ebner
2024-03-11 13:21   ` Fabian Grünbichler
2024-03-11 14:05     ` Christian Ebner
2024-03-11 15:27       ` Fabian Grünbichler
2024-03-11 15:51         ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 05/36] accessor: " Christian Ebner
2024-03-11 13:21   ` Fabian Grünbichler
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 06/36] encoder: move to stack based state tracking Christian Ebner
2024-03-11 13:21   ` Fabian Grünbichler
2024-03-11 14:12     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 07/36] encoder: add payload reference capability Christian Ebner
2024-03-11 13:21   ` Fabian Grünbichler
2024-03-11 14:15     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 08/36] encoder: add payload position capability Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 09/36] encoder: add payload advance capability Christian Ebner
2024-03-11 13:22   ` Fabian Grünbichler
2024-03-11 14:22     ` Christian Ebner
2024-03-11 15:27       ` Fabian Grünbichler
2024-03-11 15:41         ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 pxar 10/36] encoder/format: finish payload stream with marker Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 11/36] client: pxar: switch to stack based encoder state Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 12/36] client: backup: factor out extension from backup target Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 13/36] client: backup: early check for fixed index type Christian Ebner
2024-03-11 14:57   ` Fabian Grünbichler
2024-03-11 15:12     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 14/36] client: backup: split payload to dedicated stream Christian Ebner
2024-03-11 14:57   ` Fabian Grünbichler
2024-03-11 15:22     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 15/36] client: restore: read payload from dedicated index Christian Ebner
2024-03-11 14:58   ` Fabian Grünbichler
2024-03-11 15:26     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 16/36] tools: cover meta extension for pxar archives Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 17/36] restore: " Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 18/36] client: mount: make split pxar archives mountable Christian Ebner
2024-03-11 14:58   ` Fabian Grünbichler
2024-03-11 15:29     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 19/36] api: datastore: refactor getting local chunk reader Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 20/36] api: datastore: attach optional payload " Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 21/36] catalog: shell: factor out pxar fuse reader instantiation Christian Ebner
2024-03-11 14:58   ` Fabian Grünbichler
2024-03-11 15:31     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 22/36] catalog: shell: redirect payload reader for split streams Christian Ebner
2024-03-11 14:58   ` Fabian Grünbichler
2024-03-11 15:24     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 23/36] www: cover meta extension for pxar archives Christian Ebner
2024-03-11 14:58   ` Fabian Grünbichler
2024-03-11 15:31     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 24/36] index: fetch chunk form index by start/end-offset Christian Ebner
2024-03-12  8:50   ` Fabian Grünbichler
2024-03-14  8:23     ` Christian Ebner
2024-03-12 12:47   ` Dietmar Maurer
2024-03-12 12:51     ` Christian Ebner
2024-03-12 13:03       ` Dietmar Maurer
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 25/36] upload stream: impl reused chunk injector Christian Ebner
2024-03-13  9:43   ` Dietmar Maurer
2024-03-14 14:03     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 26/36] client: chunk stream: add chunk injection queues Christian Ebner
2024-03-12  9:46   ` Fabian Grünbichler
2024-03-19 10:52     ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 27/36] client: implement prepare reference method Christian Ebner
2024-03-12 10:07   ` Fabian Grünbichler
2024-03-19 11:51     ` Christian Ebner
2024-03-19 12:49       ` Fabian Grünbichler
2024-03-20  8:37         ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 28/36] client: pxar: implement store to insert chunks on caching Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 29/36] client: pxar: add previous reference to archiver Christian Ebner
2024-03-12 12:12   ` Fabian Grünbichler
2024-03-12 12:25     ` Christian Ebner
2024-03-19 12:59     ` Christian Ebner
2024-03-19 13:04       ` Fabian Grünbichler
2024-03-20  8:52         ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 30/36] client: pxar: add method for metadata comparison Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 31/36] specs: add backup detection mode specification Christian Ebner
2024-03-12 12:17   ` Fabian Grünbichler
2024-03-12 12:31     ` Christian Ebner
2024-03-20  9:28       ` Christian Ebner
2024-03-05  9:26 ` [pbs-devel] [RFC v2 proxmox-backup 32/36] pxar: caching: add look-ahead cache types Christian Ebner
2024-03-05  9:27 ` [pbs-devel] [RFC v2 proxmox-backup 33/36] client: pxar: add look-ahead caching Christian Ebner
2024-03-12 14:08   ` Fabian Grünbichler
2024-03-20 10:28     ` Christian Ebner
2024-03-05  9:27 ` Christian Ebner [this message]
2024-03-13 11:12   ` [pbs-devel] [RFC v2 proxmox-backup 34/36] fix #3174: client: pxar: enable caching and meta comparison Fabian Grünbichler
2024-03-05  9:27 ` [pbs-devel] [RFC v2 proxmox-backup 35/36] test-suite: add detection mode change benchmark Christian Ebner
2024-03-13 11:48   ` Fabian Grünbichler
2024-03-05  9:27 ` [pbs-devel] [RFC v2 proxmox-backup 36/36] test-suite: Add bin to deb, add shell completions Christian Ebner
2024-03-13 11:18   ` Fabian Grünbichler
2024-03-13 11:44 ` [pbs-devel] [RFC pxar proxmox-backup 00/36] fix #3174: improve file-level backup Fabian Grünbichler

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240305092703.126906-35-c.ebner@proxmox.com \
    --to=c.ebner@proxmox.com \
    --cc=pbs-devel@lists.proxmox.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Service provided by Proxmox Server Solutions GmbH | Privacy | Legal