From 4f09d31085fd147c59c79d42215035baf3a51c32 Mon Sep 17 00:00:00 2001
From: Dietmar Maurer <dietmar@proxmox.com>
Date: Tue, 1 Sep 2020 13:33:04 +0200
Subject: [PATCH] src/backup/verify.rs: use global hashes (instead of per
 group)

This makes verify more predictable.
---
 src/api2/admin/datastore.rs |  7 ++++---
 src/backup/verify.rs        | 28 ++++++++++++++++++++--------
 2 files changed, 24 insertions(+), 11 deletions(-)

diff --git a/src/api2/admin/datastore.rs b/src/api2/admin/datastore.rs
index f0f59ab7..0ff1631f 100644
--- a/src/api2/admin/datastore.rs
+++ b/src/api2/admin/datastore.rs
@@ -513,16 +513,17 @@ pub fn verify(
         userid,
         to_stdout,
         move |worker| {
+            let verified_chunks = Arc::new(Mutex::new(HashSet::with_capacity(1024*16)));
+            let corrupt_chunks = Arc::new(Mutex::new(HashSet::with_capacity(64)));
+
             let failed_dirs = if let Some(backup_dir) = backup_dir {
-                let verified_chunks = Arc::new(Mutex::new(HashSet::with_capacity(1024*16)));
-                let corrupt_chunks = Arc::new(Mutex::new(HashSet::with_capacity(64)));
                 let mut res = Vec::new();
                 if !verify_backup_dir(datastore, &backup_dir, verified_chunks, corrupt_chunks, worker.clone())? {
                     res.push(backup_dir.to_string());
                 }
                 res
             } else if let Some(backup_group) = backup_group {
-                verify_backup_group(datastore, &backup_group, worker.clone())?
+                verify_backup_group(datastore, &backup_group,  verified_chunks, corrupt_chunks, worker.clone())?
             } else {
                 verify_all_backups(datastore, worker.clone())?
             };
diff --git a/src/backup/verify.rs b/src/backup/verify.rs
index fd103a44..ff2894ac 100644
--- a/src/backup/verify.rs
+++ b/src/backup/verify.rs
@@ -309,7 +309,13 @@ pub fn verify_backup_dir(
 /// Returns
 /// - Ok(failed_dirs) where failed_dirs had verification errors
 /// - Err(_) if task was aborted
-pub fn verify_backup_group(datastore: Arc<DataStore>, group: &BackupGroup, worker: Arc<WorkerTask>) -> Result<Vec<String>, Error> {
+pub fn verify_backup_group(
+    datastore: Arc<DataStore>,
+    group: &BackupGroup,
+    verified_chunks: Arc<Mutex<HashSet<[u8;32]>>>,
+    corrupt_chunks: Arc<Mutex<HashSet<[u8;32]>>>,
+    worker: Arc<WorkerTask>,
+) -> Result<Vec<String>, Error> {
 
     let mut errors = Vec::new();
     let mut list = match group.list_backups(&datastore.base_path()) {
@@ -322,12 +328,6 @@ pub fn verify_backup_group(datastore: Arc<DataStore>, group: &BackupGroup, worke
 
     worker.log(format!("verify group {}:{}", datastore.name(), group));
 
-    // start with 16384 chunks (up to 65GB)
-    let verified_chunks = Arc::new(Mutex::new(HashSet::with_capacity(1024*16)));
-
-    // start with 64 chunks since we assume there are few corrupt ones
-    let corrupt_chunks = Arc::new(Mutex::new(HashSet::with_capacity(64)));
-
     BackupInfo::sort_list(&mut list, false); // newest first
     for info in list {
         if !verify_backup_dir(datastore.clone(), &info.backup_dir, verified_chunks.clone(), corrupt_chunks.clone(), worker.clone())?{
@@ -359,10 +359,22 @@ pub fn verify_all_backups(datastore: Arc<DataStore>, worker: Arc<WorkerTask>) ->
 
     list.sort_unstable();
 
+    // start with 16384 chunks (up to 65GB)
+    let verified_chunks = Arc::new(Mutex::new(HashSet::with_capacity(1024*16)));
+
+    // start with 64 chunks since we assume there are few corrupt ones
+    let corrupt_chunks = Arc::new(Mutex::new(HashSet::with_capacity(64)));
+
     worker.log(format!("verify datastore {}", datastore.name()));
 
     for group in list {
-        let mut group_errors = verify_backup_group(datastore.clone(), &group, worker.clone())?;
+        let mut group_errors = verify_backup_group(
+            datastore.clone(),
+            &group,
+            verified_chunks.clone(),
+            corrupt_chunks.clone(),
+            worker.clone(),
+        )?;
         errors.append(&mut group_errors);
     }