Martin Raiber
|
4a947874b7
|
Return error correctly when server runs out of space
|
2018-07-17 20:23:51 +02:00 |
|
Martin Raiber
|
74e899c617
|
Improve out of space error handling
(cherry picked from commit be91138dfae259f5a670ba4440647ef0b48580b9)
# Conflicts:
# urbackupserver/FileBackup.cpp
# urbackupserver/FileBackup.h
|
2018-06-01 11:50:34 +02:00 |
|
Martin
|
17d9d0737e
|
Fix build
|
2017-12-17 15:37:47 +01:00 |
|
Martin
|
8d71071ea0
|
Merge remote-tracking branch 'gh/dev' into dev
# Conflicts:
# urbackupclient/client.cpp
|
2017-03-03 17:33:47 +01:00 |
|
Nick Glencross
|
3f4e8efb9b
|
Tweak 'successfull' -> 'successful'
|
2017-02-28 09:02:15 +04:00 |
|
Martin Raiber
|
9076cec45b
|
Invalidate last patches when reconnecting OFB pipe
(cherry picked from commit ab676496f1)
|
2017-02-12 22:28:54 +01:00 |
|
Martin
|
c85129b490
|
More detailed error message when receiving unknown packet id
|
2017-01-31 15:50:26 +01:00 |
|
Martin
|
160e7c2549
|
Increase robustness of OFB block load
|
2017-01-31 15:46:12 +01:00 |
|
Martin
|
28d2324e16
|
Lock with correct mutex
|
2017-01-16 00:06:27 +01:00 |
|
Martin
|
36594622f8
|
Fix bug with sparse extent resetting during chunked file transfer reconnect
(cherry picked from commit 569580d3c1d084702e39c1a191f87fc02f263073)
|
2017-01-12 13:59:43 +01:00 |
|
Martin Raiber
|
8705688342
|
Clear initial bytes after reconnect
|
2016-12-28 19:00:41 +01:00 |
|
Martin
|
2af0c785d6
|
Fix reconnect after chunk/file request timeout
|
2016-12-22 02:33:44 +01:00 |
|
Martin
|
0e2420cf74
|
Handle send failure in file client gracefully and not by accidentally queuing two same chunks
|
2016-12-20 21:14:13 +01:00 |
|
Martin
|
cab42dedcb
|
Set FD_CLOEXEC via fcntl if SOCK_CLOEXEC is not available
|
2016-12-08 00:49:47 +01:00 |
|
Martin
|
a5fcebe5dc
|
Set FD_CLOEXEC via fcntl if SOCK_CLOEXEC is not available
|
2016-12-08 00:29:16 +01:00 |
|
Martin
|
601d901c91
|
Fix rare issue with out-of-band block transfer
|
2016-12-04 16:26:50 +01:00 |
|
Martin
|
0632b25586
|
Remove only read part from buffer
|
2016-10-19 17:42:46 +02:00 |
|
Martin
|
740961c156
|
Catch instances where a client with empty name is accidentally added
(cherry picked from commit 9dba295c7378ff5a81ccf2b5c403c6d4d80dba7c)
|
2016-09-28 12:55:15 +02:00 |
|
Martin
|
e6deace398
|
Preallocate space for full image if possible
|
2016-08-12 02:14:03 +02:00 |
|
Martin
|
fc601cd475
|
Do not cleanup on writing failures during file metadata download
|
2016-08-07 17:36:04 +02:00 |
|
Martin
|
4c8a54b34d
|
Do not switch over to next file if in accumulation state and file size is zero
|
2016-06-27 00:23:36 +02:00 |
|
Martin
|
8636556999
|
Log error codes upon receiving them
|
2016-06-10 15:26:27 +02:00 |
|
Martin
|
60a167aa85
|
Set error codes correctly when download is queued
|
2016-06-10 03:12:03 +02:00 |
|
Martin
|
b8a4bafde9
|
Set error codes correctly when download is queued
|
2016-06-10 03:11:18 +02:00 |
|
Martin
|
6ead30936d
|
Improved read error handling during file backup
|
2016-06-09 23:33:32 +02:00 |
|
Martin
|
22b6766da9
|
Further improve read error handling
(cherry picked from commit 18945955a5a14ac763d71bbafef42b8fbcae56c5)
|
2016-06-09 22:17:30 +02:00 |
|
Martin
|
b9a14e5e4b
|
Limit number of reconnect tries for chunked transfer to 50
|
2016-06-08 18:03:03 +02:00 |
|
Martin
|
61f2ff400b
|
Lower reconnection timeout if shutting down file client
|
2016-05-30 16:57:18 +02:00 |
|
Martin
|
21a51b6d39
|
Renamed ERR_FILE_DOESNT_EXIST to ERR_CANNOT_OPEN_FILE. Do not stop backup if this error occurs during patch file download
|
2016-05-28 21:31:42 +02:00 |
|
Martin Raiber
|
054cb789bc
|
Fix issues discovered by code analysis
|
2016-05-17 21:08:41 +02:00 |
|
Martin Raiber
|
cea0771256
|
Update output file size only when writing out patch and not earlier
|
2016-05-15 21:47:03 +02:00 |
|
Martin Raiber
|
60e0d28f16
|
Do not reconnect from file queuing because this causes double frees
|
2016-05-15 20:22:06 +02:00 |
|
Martin
|
d478e830ae
|
Reconnect if file request fails and retry
|
2016-05-13 21:37:37 +02:00 |
|
Martin
|
7703db1323
|
Resize hash output to correct size if it is lower than what was already output
|
2016-05-12 17:21:53 +02:00 |
|
Martin
|
489b21fcd1
|
Set resume flag only if something was already transferred
|
2016-05-11 19:00:28 +02:00 |
|
Martin
|
40d970214b
|
Do not re-created script and meta-data transfers when client is restarted and server reconnects
|
2016-05-10 19:03:35 +02:00 |
|
Martin
|
19d79e8e8f
|
Improved handling of case where snapshot removal is delayed because of still ongoing meta-data transfers
|
2016-05-08 16:28:38 +02:00 |
|
Martin
|
d485e1fe8e
|
Restore fixes
|
2016-05-06 23:42:21 +02:00 |
|
Martin
|
92141c5ea8
|
Log about download with sparse extents
|
2016-05-05 16:04:26 +02:00 |
|
Martin
|
fa1f3f5715
|
Load whole block regardless of block hash if reading for hash fails
|
2016-05-05 00:22:10 +02:00 |
|
Martin
|
df7f2a95fe
|
Fix restore reconnect segfault
|
2016-05-02 18:05:12 +02:00 |
|
Martin
|
c5b1f2f525
|
Fix bug with zero size file restore
|
2016-05-02 17:18:46 +02:00 |
|
Martin
|
54a8201938
|
Fix restore segfault
|
2016-05-02 15:13:15 +02:00 |
|
Martin
|
b81ecb236a
|
Reset curr_sparse_extent to prevent segfault
|
2016-05-01 18:19:12 +02:00 |
|
Martin
|
4d9a889b0c
|
Correctly handle zero length chunks again
(cherry picked from commit 55559e8c927a365940574fa2ec1c193fa863d467)
|
2016-04-29 02:26:55 +02:00 |
|
Martin
|
d715248f55
|
Correctly handle zero length chunks again
(cherry picked from commit d96ec4e0a8a8529c89f800d19f387bc8103bec70)
|
2016-04-29 01:47:37 +02:00 |
|
Martin
|
84d48f8fe3
|
Log OS errors in block differences transfer
|
2016-04-26 23:47:21 +02:00 |
|
Martin
|
ddb4def151
|
Do not switch state for zero length chunk
(cherry picked from commit f565154d5691d5422269cc34d0a52a0d2c813068)
|
2016-04-20 13:16:59 +02:00 |
|
Martin
|
b3c5cfd684
|
Reset extent iterator
|
2016-03-30 16:58:54 +02:00 |
|
Martin
|
b070462c29
|
Fixed hang with stuck socket
|
2016-02-21 16:18:20 +01:00 |
|