Commit Graph

727 Commits

Author SHA1 Message Date
Wolfgang Bumiller
dcd15d10bb backup/index: remove IndexIterator
removed the need for it, to avoid copies

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-03-01 13:29:36 +01:00
Dietmar Maurer
541b908ea4 src/backup/datastore.rs: use DateTime<Utc> instead of unix epoch 2019-02-28 17:03:48 +01:00
Wolfgang Bumiller
5de2bced2d datastore: add generic open_index
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-28 15:26:40 +01:00
Wolfgang Bumiller
7bc1d72778 add IndexFile trait
We want to be able to iterate through digests of index
files, but without always having to distinguish between
dynamic and fixed types, so add a trait we can use as a
trait object.

Unfortunately the iterator needs to yield copies as
iterators cannot yield values with life times when
represented as trait objects (Box<dyn Iterator<Item = ?>>)

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-28 15:26:40 +01:00
Wolfgang Bumiller
5be4065baf make FixedIndexReader Send and update comments
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-28 15:26:40 +01:00
Wolfgang Bumiller
fded74d0c7 style: BufferedReader -> BufferedRead
Single-method traits usually use the same name as their
method and aren't usually the 'noun' (which is usually an
implementation of them instead).

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-28 15:26:40 +01:00
Dietmar Maurer
8f57971714 src/api2/admin/datastore.rs: implement prune 2019-02-28 12:51:27 +01:00
Wolfgang Bumiller
8731e40a7f backup/datastore: create_backup_dir: show if path existed
To enable asserting the creation of a new backup rather than
adding to an existing one.

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-27 10:15:36 +01:00
Wolfgang Bumiller
f661f37427 backup/dynamic_index: split out add_chunk
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-25 13:50:45 +01:00
Wolfgang Bumiller
9ac6ec868a backup/chunk_store: split insert_chunk
The protocol handler will receive chunk data plus a hash
pre-calculated by the client. It will verify the hash before
sending it up to the datastore in order to respond to the
client with an error on a mismatch, so there's no need to
recalculate the hash another time.

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-25 13:44:29 +01:00
Wolfgang Bumiller
3a50ddd0c9 backup/chunk_store: mark ChunkIterator as fused
Ranges are FusedIterators and it can be useful to propagate
that property to users.

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-25 13:44:29 +01:00
Wolfgang Bumiller
eff25ecaf2 backup/chunk_store: make percentage output optional
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-25 13:44:29 +01:00
Wolfgang Bumiller
d59397e60e backup/datastore: expose chunk_store.get_chunk_iterator
will be used by the protocol handler to iterate over
existing chunks

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-25 13:44:29 +01:00
Dietmar Maurer
8eb1203bda src/backup/chunk_stat.rs: log write speed 2019-02-25 13:21:00 +01:00
Dietmar Maurer
7e3365554e src/backup/chunk_stat.rs: new struct to track chunk statistics 2019-02-25 12:52:10 +01:00
Dietmar Maurer
798f7fa065 src/backup/chunk_store.rs: return info about compressed chunk size
So that we can generate better statistics ..
2019-02-25 11:36:05 +01:00
Dietmar Maurer
78216a5ab1 src/backup/chunk_store.rs: use lz4 for compression
And depend on lz4 lib now.
2019-02-25 10:00:28 +01:00
Dietmar Maurer
f252ca654c src/backup/chunker.rs: simplify break test
To make it a little bit faster ...
2019-02-22 13:02:11 +01:00
Dietmar Maurer
61834c1cbf src/backup/chunker.rs: another try to speedup the chunker 2019-02-22 11:15:10 +01:00
Dietmar Maurer
8c75372b79 src/backup/datastore.rs: list_backups() - include list of index files 2019-02-20 13:37:44 +01:00
Dietmar Maurer
fa4ebf0fce src/backup/chunker.rs: simplify code 2019-02-20 11:17:19 +01:00
Dietmar Maurer
1b9c959c32 src/backup/chunker.rs: change window size back to 48
Using 64 is a bad idea, because it always computes hash 0 for
constant data streams.
2019-02-20 11:01:09 +01:00
Dietmar Maurer
247cdbce72 src/api2/admin/datastore/catar.rs: allow to configure chunk-size 2019-02-19 15:19:12 +01:00
Dietmar Maurer
c584aa21fd src/backup/chunker.rs: start() - correctly store hash in self.h 2019-02-19 12:40:12 +01:00
Dietmar Maurer
cd1598302a src/backup/chunker.rs: use 64 byte chunker window size
This results in faster code.
2019-02-19 12:27:37 +01:00
Dietmar Maurer
9a3e869cf5 src/backup/chunker.rs: simplify buzzhash start code 2019-02-19 12:17:17 +01:00
Dietmar Maurer
4d2881dd4c src/backup/chunker.rs: correctly compute window index 'idx'
And add a simple test.
2019-02-18 16:59:23 +01:00
Dietmar Maurer
f0a6112451 backup/datastore.rs: improve error reporting 2019-02-15 14:33:03 +01:00
Dietmar Maurer
af6f80d3db Revert "backup/datastore: fn open is a private implementation detail"
This reverts commit 3dacd9a660.

I currently use this in some test ...
2019-02-14 13:13:49 +01:00
Wolfgang Bumiller
62f2422f6a backup/chunk_store: verify chunk file names
get_chunk_iterator() should skip over files which aren't an
actual chunk

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-14 12:37:31 +01:00
Wolfgang Bumiller
3dacd9a660 backup/datastore: fn open is a private implementation detail
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-14 12:37:31 +01:00
Wolfgang Bumiller
9739aca41f backup/chunk_store: open base dir in get_chunk_iterator
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-14 12:37:31 +01:00
Wolfgang Bumiller
c7f481b642 backup: replace ChunkIterator
Instead we just build the iterator using our
tools::fs::read_subdir iterator as follows:
Use the index range (0..0x10000) as an iterator, map() it
to yield a subdirectory iterator for each index, then
flatten it.

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-14 10:09:59 +01:00
Dietmar Maurer
e5064ba607 simplify backup lib structure (pub use xxx:*), improve doc 2019-02-12 14:13:31 +01:00
Dietmar Maurer
93d5d77952 rename ArchiveIndex to DynamicIndex
also changed the file extension from .aidx to .didx
2019-02-12 12:05:33 +01:00
Dietmar Maurer
91a905b6dd rename ImageIndex to FixedIndex
also changed the file extension fron .iidx to .fidx
2019-02-12 11:50:45 +01:00
Wolfgang Bumiller
82bc0ad40c backup: fixup warnings and readd lost file type check
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-12 10:43:43 +01:00
Wolfgang Bumiller
9349d2a1d2 backup: cleanup long lines
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-12 10:39:11 +01:00
Wolfgang Bumiller
fdd71f5286 backup: use ChunkIterator for sweep_unused_chunks
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-12 10:39:11 +01:00
Wolfgang Bumiller
c9e7f4dc8b backup: add optional progress callback to ChunkIterator
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-12 10:39:11 +01:00
Wolfgang Bumiller
4070096c6a backup: add ChunkIterator
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-12 10:39:11 +01:00
Wolfgang Bumiller
e0a5d1ca2f tests: chunk_store: cleanup .testdir
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2019-02-01 10:41:54 +01:00
Dietmar Maurer
9f49fe1d5d avoid compiler warnings 2019-01-30 18:25:37 +01:00
Dietmar Maurer
2296860027 tools.rs: move digest_to_hex() into tools 2019-01-25 10:58:28 +01:00
Dietmar Maurer
332dcc2277 backup/chunk_store.rs: fix test case - use absolute path 2019-01-25 10:14:25 +01:00
Dietmar Maurer
784252dbc5 backup/datastore.rs: use single lazy_static block 2019-01-24 14:58:47 +01:00
Dietmar Maurer
7ca80246ea backup/datastore.rs: use DateTime<Utc> to store backup time 2019-01-21 10:51:52 +01:00
Dietmar Maurer
be0084b001 white space fix 2019-01-21 10:32:07 +01:00
Dietmar Maurer
e25736b4c3 backup/datastore.rs: implement list_backups 2019-01-20 18:02:27 +01:00
Dietmar Maurer
68469eebb4 backup/chunk_store.rs: verify base path is absolute 2019-01-20 16:49:22 +01:00
Dietmar Maurer
0b05fd5830 api3/admin/datastore/catar.rs: implement download 2019-01-20 09:39:32 +01:00
Dietmar Maurer
6a4c091616 api3/admin/datastore/catar.rs: start impl. download 2019-01-19 17:27:02 +01:00
Dietmar Maurer
728797d0c1 reduce compiler warnings 2019-01-18 16:50:15 +01:00
Dietmar Maurer
95cea65b04 backup/datastore.rs: list all index files using walkdir crate 2019-01-18 12:24:58 +01:00
Dietmar Maurer
ff3d3100d4 backup/datastore.rs: try to create useful directory layout
store/type/ID/TIME/xxx.yyy
2019-01-18 12:01:37 +01:00
Dietmar Maurer
150f1bd8f6 backup/datastore.rs: use Arc for chunk_store references 2019-01-15 12:36:16 +01:00
Dietmar Maurer
1629d2ad7b api3/admin/datastore/upload_catar.rs: implement upload future 2019-01-15 11:38:26 +01:00
Dietmar Maurer
ddbdf80d51 avoid compiler warnings 2019-01-11 08:41:33 +01:00
Dietmar Maurer
373ef4a504 backup/archive_index.rs: correctly check for errors 2019-01-10 11:19:54 +01:00
Dietmar Maurer
4624fe29cc impl Read for BufferedArchiveReader 2019-01-06 10:04:45 +01:00
Dietmar Maurer
343370504b impl Seek for BufferedArchiveReader 2019-01-06 09:35:39 +01:00
Dietmar Maurer
318564ac03 buffered_read: return empty buffer on eof 2019-01-06 09:17:28 +01:00
Dietmar Maurer
0a72e26704 define and use BufferedReader trait 2019-01-05 17:28:20 +01:00
Dietmar Maurer
39c6bd86cc backup/archive_index.rs: implement BufferedArchiveReader
Implement relativly fast random read using binary search.
2019-01-05 14:47:56 +01:00
Dietmar Maurer
060c4811a0 backup/chunk_store.rs: resize buffer by powers of two 2019-01-05 11:34:49 +01:00
Dietmar Maurer
df9973e8f5 backup/chunk_store.rs: avoid allocation in read_chunk 2019-01-04 17:16:56 +01:00
Dietmar Maurer
7394ca3e95 white space cleanups 2019-01-04 12:51:43 +01:00
Dietmar Maurer
96df2fb44b backup/chunk_store.rs: implement read_chunk() 2019-01-04 12:50:54 +01:00
Dietmar Maurer
bc616633d5 backup/chunk_store.rs: only use one directory level 2019-01-04 10:35:22 +01:00
Dietmar Maurer
b663789bdb backup/archive_index.rs: pass correct size to unmap 2019-01-04 09:28:41 +01:00
Dietmar Maurer
48d0d35688 backup/image_index.rs: use little endian 2019-01-04 08:45:45 +01:00
Dietmar Maurer
050f8a6355 chunker.rs: make sure chunk_size_avg is a power of two 2019-01-03 15:10:54 +01:00
Dietmar Maurer
ea4ea34baf improve chunker speed (avoid modulo) 2019-01-03 14:35:19 +01:00
Dietmar Maurer
0b8e75ed16 image_index.rs: verify file size 2019-01-02 18:14:02 +01:00
Dietmar Maurer
77703d95aa implement garbage collection for .aidx files 2019-01-02 14:27:04 +01:00
Dietmar Maurer
a360f6fa2d image_index.rs: verify header magic/version 2019-01-02 13:13:13 +01:00
Dietmar Maurer
44b3f62b42 fix typo 2019-01-02 12:56:04 +01:00
Dietmar Maurer
5032b57b44 archive_index.rs: correctly write .aidx file 2019-01-02 12:54:40 +01:00
Dietmar Maurer
580dc84bac image_index.rs: print stats 2019-01-02 12:53:49 +01:00
Dietmar Maurer
5e7a09be0d src/backup/archive_index.rs: use close() instead of flush()
Also pass a reference to the encoder.
2019-01-02 11:02:56 +01:00
Dietmar Maurer
94a882e900 src/backup/archive_index.rs: implement flush() 2018-12-31 18:01:07 +01:00
Dietmar Maurer
0433db1949 src/backup/archive_index.rs: first try 2018-12-31 17:30:08 +01:00
Dietmar Maurer
cbdd8c54ae create backup mod in backup.rs, improve docu 2018-12-31 16:08:04 +01:00
Dietmar Maurer
cb4412b18e move chunker.rs to backup/ 2018-12-31 15:46:16 +01:00
Dietmar Maurer
bcd879cfb3 backup-client: allow to specify directories (prepare for catar backup) 2018-12-27 10:11:11 +01:00
Dietmar Maurer
a198d74fc0 cleanup: reduce compiler warnings 2018-12-25 13:29:27 +01:00
Dietmar Maurer
991de6007b remove outdated comment 2018-12-25 13:24:50 +01:00
Dietmar Maurer
7ee2aa1b94 touch_chunk: use libc::lutimensat 2018-12-25 12:27:25 +01:00
Dietmar Maurer
2c32fdde86 move lookup_datastore() to backup/datastore.rs 2018-12-22 17:37:25 +01:00
Dietmar Maurer
64e53b2835 gather usage statistics during garbage collection 2018-12-22 16:58:16 +01:00
Dietmar Maurer
176e4af964 sweep_used_chunks: print percentage 2018-12-22 15:39:05 +01:00
Dietmar Maurer
7b2b40a893 use openssl for faster hashing 2018-12-22 14:31:59 +01:00
Dietmar Maurer
e95950e40a chunk_store: reduce number of directories
Else, scans everything takes too long ...
2018-12-22 14:04:05 +01:00
Dietmar Maurer
03e4753d8e fix mutability for chunk store 2018-12-21 12:15:26 +01:00
Dietmar Maurer
d62e6e2264 backup-client: use 1M chunks, make chunk_size configurable 2018-12-21 08:36:57 +01:00
Dietmar Maurer
f7dd683b46 fix tests 2018-12-19 13:54:22 +01:00
Dietmar Maurer
277fc5a3ce try to improve error messages 2018-12-19 13:40:26 +01:00
Dietmar Maurer
eae8aa3aa9 cleanups: avoid compiler warnings 2018-12-19 12:49:23 +01:00
Dietmar Maurer
6c20a13d3c cleanup - avoid too much indentation 2018-12-19 12:40:26 +01:00
Dietmar Maurer
1c43c56b06 cleanup 2018-12-19 12:30:37 +01:00
Dietmar Maurer
2bf5f6b220 ChunkStore::sweep_used_chunks - use openat to speedup directory walks 2018-12-19 12:20:01 +01:00
Wolfgang Bumiller
1628a4c731 use timers with a signal for file locking
* rename lock_file -> open_file_locked,
* add lock_file as a function working on already-opened
  files
* change timeout types to std::time::Duration

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2018-12-19 11:23:00 +01:00
Dietmar Maurer
365bb90f17 move lock_file to tools.rs 2018-12-19 10:02:24 +01:00
Dietmar Maurer
08481a0b60 impl sweep_used_chunks, first try 2018-12-19 09:51:33 +01:00
Dietmar Maurer
6ea3a0b7fc move sweep_used_chunks to chunk_store 2018-12-18 11:18:55 +01:00
Dietmar Maurer
3d5c11e5d2 start impl garbage collection 2018-12-18 11:06:03 +01:00
Dietmar Maurer
529de6c7a3 start implementing DataStore 2018-12-17 13:00:39 +01:00
Dietmar Maurer
4818c8b6f7 start implementing ImageIndexReader 2018-12-16 14:44:44 +01:00
Dietmar Maurer
af3e7d7551 fix chunk store file paths 2018-12-16 13:52:16 +01:00
Dietmar Maurer
0cd9d4206f improve error messages 2018-12-16 13:43:19 +01:00
Dietmar Maurer
4fbb72a8b4 image_index.rs: implement atomic write/rename 2018-12-16 13:39:21 +01:00
Dietmar Maurer
798881a68a rebane ImageIndex to ImageIndexWriter 2018-12-16 11:51:16 +01:00
Dietmar Maurer
5e5b7f1c2d use u64 to store time 2018-12-16 11:48:03 +01:00
Dietmar Maurer
d13e3745b3 image_index.rs: write header 2018-12-15 17:05:49 +01:00
Dietmar Maurer
606ce64bb0 image_index.rs:write idx of chunks (use mmap) 2018-12-15 14:51:05 +01:00
Dietmar Maurer
5ba6968922 add comment to make intensions more clear 2018-12-12 11:21:00 +01:00
Dietmar Maurer
ce55dbbcc6 simplify code 2018-12-09 09:42:17 +01:00
Dietmar Maurer
461e62fcdf remove some compiler warnings 2018-12-08 14:55:54 +01:00
Dietmar Maurer
391a2e43ff chunk_store.insert_chunk: return information about existing chunks 2018-12-08 11:25:11 +01:00
Dietmar Maurer
12bb93b3ad code cleanup 2018-12-08 11:20:02 +01:00
Dietmar Maurer
022fb421b2 lock with timeout
clumsy, but I have node idea how to make it better
2018-12-08 10:57:09 +01:00
Dietmar Maurer
2989f6bf20 improve error messages 2018-12-08 08:38:42 +01:00
Dietmar Maurer
b8d4766a4d code cleanup 2018-12-08 08:21:20 +01:00
Dietmar Maurer
45773720ab depend on nix, use flock
Signed-off-by: Dietmar Maurer <dietmar@proxmox.com>
2018-12-07 18:14:07 +01:00
Dietmar Maurer
c5d82e5f34 chunk_store: use mutex 2018-12-07 16:12:45 +01:00
Dietmar Maurer
128b37feea impl insert_chunk 2018-12-07 14:44:56 +01:00
Dietmar Maurer
35cf5daa53 start experiments with simple chunk store 2018-12-07 10:51:53 +01:00