summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* ctdb-recoverd: For persistent databases a sequence number of 0 is validMartin Schwenke2013-11-271-2/+3
| | | | | | | | Otherwise recovery ends up done by RSN when it is unnecessary. Signed-off-by: Martin Schwenke <martin@meltin.net> Pair-programmed-with: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-locking: Use vfork instead of fork to exec helpersAmitay Isaacs2013-11-271-2/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There is a significant overhead using fork() over vfork(), specially when the child process execs a helper. The overhead is in memory space and time. # strace -c ./test_fork 1024 200 count=1024, size=204800, total=200M failed fork=0 time for fork() = 4879.597000 us % time seconds usecs/call calls errors syscall ------ ----------- ----------- --------- --------- ---------------- 100.00 4.543321 3304 1375 375 clone 0.00 0.000071 0 1033 mmap 0.00 0.000000 0 1 read 0.00 0.000000 0 3 write 0.00 0.000000 0 2 open 0.00 0.000000 0 2 close 0.00 0.000000 0 3 fstat 0.00 0.000000 0 3 mprotect 0.00 0.000000 0 1 munmap 0.00 0.000000 0 3 brk 0.00 0.000000 0 1 1 access 0.00 0.000000 0 1 execve 0.00 0.000000 0 1 arch_prctl ------ ----------- ----------- --------- --------- ---------------- 100.00 4.543392 2429 376 total # strace -c ./test_vfork 1024 200 count=1024, size=204800, total=200M failed fork=0 time for fork() = 82.041000 us % time seconds usecs/call calls errors syscall ------ ----------- ----------- --------- --------- ---------------- 96.47 0.001204 1 1000 vfork 3.53 0.000044 0 1033 mmap 0.00 0.000000 0 1 read 0.00 0.000000 0 3 write 0.00 0.000000 0 2 open 0.00 0.000000 0 2 close 0.00 0.000000 0 3 fstat 0.00 0.000000 0 3 mprotect 0.00 0.000000 0 1 munmap 0.00 0.000000 0 3 brk 0.00 0.000000 0 1 1 access 0.00 0.000000 0 1 execve 0.00 0.000000 0 1 arch_prctl ------ ----------- ----------- --------- --------- ---------------- 100.00 0.001248 2054 1 total Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-common: Refactor code to keep track of child processesAmitay Isaacs2013-11-272-8/+15
| | | | | | | This code can then be used to track child processes created with vfork(). Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-scripts: Run a single instance of debug_locks.sh at a give timeAmitay Isaacs2013-11-271-24/+34
| | | | | | | | | | This prevents spamming of logs if multiple lock requests are waiting and keep timing out. Also, improve the logging format with separators. Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-locking: Update current lock statistics when lock is scheduledAmitay Isaacs2013-11-271-2/+2
| | | | | | | | | When a child process is created for a lock request, the current locks statistics should be updated immediately. This will provide accurate information on number of active lock requests. Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-locking: Do not merge multiple lock requests to avoid unfair schedulingAmitay Isaacs2013-11-271-1/+6
| | | | | Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-locking: Implement active lock requests limit per databaseAmitay Isaacs2013-11-272-9/+13
| | | | | | | | | | | | This limit was currently a global limit and not per database. This prevents any database freeze lock requests from getting scheduled if the global limit was reached. Only individual record requests should be limited and database freeze requests should always get scheduled. Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-scripts: Rewrite statd-callout to avoid 10 minute lagMartin Schwenke2013-11-273-113/+96
| | | | | | | | | This is naive and assumes no performance problems when updating persistent DBs. It also does no error handling. Signed-off-by: Martin Schwenke <martin@meltin.net> Pair-programmed-with: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-client: Treat empty __db_sequence_number__ record as 0Amitay Isaacs2013-11-271-0/+5
| | | | | | | | | This fixes the issue of transaction commit failing due to an empty __db_sequence_number__ record in persistent database left by previous cancelled transaction. Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-doc: Update ctdb.1 - primarily to add pdelete/pfetch/pstore/ptransMartin Schwenke2013-11-271-129/+195
| | | | | | | | | | | | | | | | | | | Also: * More <refentryinfo> above <refmeta> to make the XML valid. * Describe DB argument in introduction and use it for database commands. * Remove unnecessary format="linespecific" from <screen> tags, since it will not be allowed in DocBook 5.0. * Sort the items in "INTERNAL COMMANDS". * Update/simplify some command descriptions. Signed-off-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-tools/ctdb: New ptrans commandMartin Schwenke2013-11-273-0/+278
| | | | | | | | Also add test. Signed-off-by: Martin Schwenke <martin@meltin.net> Pair-programmed-with: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-onnode: New -i option to stop stdin from being closedMartin Schwenke2013-11-272-5/+22
| | | | | | | | | | This can be useful for piping data to onnode in certain circumstances. There are now also enough command-line options that they should definitely be alphabetically ordered. Signed-off-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-tests/integration: try_command_on_node() shouldn't lose onnode optionsMartin Schwenke2013-11-271-1/+1
| | | | | | | | Currently it only passes the last (non -v) option seen. It should pass them all. Signed-off-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-recoverd: Fix backward compatibility for CTDB_SRVID_TAKEOVER_RUNMartin Schwenke2013-11-273-20/+42
| | | | | | | | | When running a mixed version cluster, compatibility with older versions was was broken during recent refactorisation. Signed-off-by: Martin Schwenke <martin@meltin.net> Pair-programmed-with: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-scripts: debug_locks.sh should use configuration to find TDB locationMartin Schwenke2013-11-271-2/+10
| | | | | | | | That is, don't use fixed paths. Signed-off-by: Martin Schwenke <martin@meltin.net> Pair-programmed-with: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-recoverd: A node refuses to play against itselfMartin Schwenke2013-11-271-0/+5
| | | | | | Signed-off-by: Martin Schwenke <martin@meltin.net> Pair-programmed-with: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-recoverd: Remove duplicate code to update flags during recoveryMartin Schwenke2013-11-271-17/+0
| | | | | | | | This also happens earlier in do_recovery() and the nodemap is not updated after that, so this update is redundant. Signed-off-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-build: Update to latest upstream config.guessMartin Schwenke2013-11-271-63/+86
| | | | | Signed-off-by: Martin Schwenke <martin@meltin.net> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-tools/ctdb: Fix db commands when dbid is given instead of nameAmitay Isaacs2013-11-271-32/+31
| | | | | Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-tests: CTDB tool should always be invoked as $CTDB instad of ctdbAmitay Isaacs2013-11-277-92/+92
| | | | | | | | $CTDB_TEST_WRAPPER is required only to run test functions or test binaries on remote nodes. For running ctdb command, $CTDB is sufficient. Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-tests: No need to run onnode in parallel for single nodeAmitay Isaacs2013-11-273-6/+6
| | | | | Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* ctdb-tests: Remove -q option to try_command_on_nodeAmitay Isaacs2013-11-2712-99/+99
| | | | | | | This option is always passed to onnode by default. Signed-off-by: Amitay Isaacs <amitay@gmail.com> Reviewed-by: Michael Adam <obnox@samba.org>
* build: check linux/fs.h for compression flagsDavid Disseldorp2013-11-271-1/+2
| | | | | | | | | | | | | | Needed for old distributions (e.g. SLES8), which provide a linux/fs.h without the FS_IOC_GETFLAGS/FS_COMPR_FL defines needed for the btrfs VFS module. Reported-by: Andrew Bartlett <abartlet@samba.org> Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Autobuild-User(master): Andrew Bartlett <abartlet@samba.org> Autobuild-Date(master): Wed Nov 27 18:44:14 CET 2013 on sn-devel-104
* s3:smb2_server: avoid calling set_current_user_info() for each requestStefan Metzmacher2013-11-273-7/+9
| | | | | | | | Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: David Disseldorp <ddiss@samba.org> Autobuild-User(master): Stefan Metzmacher <metze@samba.org> Autobuild-Date(master): Wed Nov 27 16:31:44 CET 2013 on sn-devel-104
* s3:smb2_server: generate a header blob for the sendfile pathStefan Metzmacher2013-11-273-5/+40
| | | | | | | | | | We need to pass the NBT header, SMB2 header and SMB2 Read header as header blob to SMB_VFS_SENDFILE(). This allows the usage of MSG_SEND or other tricks to avoid multiple TCP packets on the wire. Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: David Disseldorp <ddiss@samba.org>
* s3:smb2_server: allocate smbd_smb2_request on talloc_tos()Stefan Metzmacher2013-11-271-1/+3
| | | | | | | | This matches the behavior for smb1 requests and avoids an additional malloc() per request. Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: David Disseldorp <ddiss@samba.org>
* s3:smb2_server: use tevent_req_notify_callback() in ↵Stefan Metzmacher2013-11-271-0/+7
| | | | | | | | | | | smbd_smb2_request_pending_queue() If the request is already done we can avoid one iteration of tevent_loop_once(), which means we avoids one talloc_stackframe_pool/talloc_free pair. Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: David Disseldorp <ddiss@samba.org>
* s3:smb2_server: for performance reasons we use tevent_fd and readv/writev ↵Stefan Metzmacher2013-11-272-438/+402
| | | | | | | | | | | | | | | | | | | | | directly Going via tevent_req_create/talloc_free at multiple layer costs too much cpu cycles per request. I tested downloading a 16GB (sparse) file with smbclient -b1 -mNT1, and -mSMB2_02. Using smb2 max read = 64512, which means smb1 and smb2 will use the same read size. I build with -O3 -g and compared the results with valgrind --tool=callgrind. With -mNT1 the server uses about 2.000.000.000 cpu cycles. This patch reduces the userspace cpu cycles for -mSMB2_02 from about ~ 8.000.000.000 down to ~ 4.000.000.000. Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: David Disseldorp <ddiss@samba.org>
* s3:smb2_server: fix drain_socket error handlingStefan Metzmacher2013-11-271-3/+17
| | | | | | | | smbd_smb2_request_error_ex() should return NTSTATUS and the caller will terminate the connection. Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: David Disseldorp <ddiss@samba.org>
* smbd: Fix a typoVolker Lendecke2013-11-271-1/+1
| | | | | Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
* smbd: Remove a "set but unused" variableVolker Lendecke2013-11-271-1/+0
| | | | | | | | | | This is should have been in a793ac0 Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org> Autobuild-User(master): Stefan Metzmacher <metze@samba.org> Autobuild-Date(master): Wed Nov 27 13:30:48 CET 2013 on sn-devel-104
* torture3: Reproducer for bug 10284Volker Lendecke2013-11-264-0/+89
| | | | | | | | Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org> Autobuild-User(master): Volker Lendecke <vl@samba.org> Autobuild-Date(master): Tue Nov 26 22:53:04 CET 2013 on sn-devel-104
* smbd: Fix bug 10284Volker Lendecke2013-11-261-69/+31
| | | | | | | | | | | | | | | | | | If we msg_read_send on a nonempty channel, we create one tevent_immediate. If we directly receive another message and from within the msg_read_send's tevent_req callback we immediately do another msg_read_send, we end up with two tevent_immediate events for msg_channel_trigger with just one incoming message. Test to follow. This patch simplifies msg_channel.c by removing the explicit immediate events. Instead, it relies on the implicit immediate event available via tevent_req_defer_callback. For messages received from tdb with a msg_read_send req pending, we directly finish that request without putting the message on the queue. Bug: https://bugzilla.samba.org/show_bug.cgi?id=10284 Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org>
* smbd: Pull mtime handling into open_file_ntcreateVolker Lendecke2013-11-234-38/+23
| | | | | | | | | | | | | | This is a performance improvement for heavily contended files, in particular in a cluster. The separate call to get_file_infos makes us pull the locking.tdb record twice per open. For a contended file this can be a performance penalty, this gets the # of record accesses for the open/close cycle down from 3 to 2. Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org> Autobuild-User(master): Jeremy Allison <jra@samba.org> Autobuild-Date(master): Sat Nov 23 00:40:49 CET 2013 on sn-devel-104
* smbd: Point reply_open at the correct stat structVolker Lendecke2013-11-221-1/+4
| | | | | | | | Not sure if anybody will ever notice this these days, but the same is done in the createfile calls. Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* smbd: Factor out get_share_mode_write_time of get_file_infosVolker Lendecke2013-11-222-8/+12
| | | | | Signed-off-by: Volker Lendecke <vl@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* torture: test get/set compression ioctl permissionsDavid Disseldorp2013-11-221-0/+128
| | | | | | | | | | | | | Windows Server 2012[r2] exhibits some strange behaviour with regard to handling the compression fsctls. [READ/WRITE]_ATTR permissions are not required for the corresponding get/set compression ioctls. WRITE_DATA is required for set compression. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org> Autobuild-User(master): Jeremy Allison <jra@samba.org> Autobuild-Date(master): Fri Nov 22 19:57:48 CET 2013 on sn-devel-104
* torture: split open from test_setup_create_fillDavid Disseldorp2013-11-221-12/+32
| | | | | Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* smbd/open: disable compression with FILE_NO_COMPRESSIONDavid Disseldorp2013-11-221-0/+12
| | | | | | | | Files and directories created with FILE_NO_COMPRESSION should not inherit the compression attribute from their parent directory. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* s3-smbd: support FILE_ATTRIBUTE_COMPRESSEDDavid Disseldorp2013-11-221-0/+47
| | | | | | | | | | The FILE_ATTRIBUTE_COMPRESSED flag is computed based on whether the filesystem exposes the FILE_FILE_COMPRESSION capability, and whether SMB_VFS_GET_COMPRESSION() reports that the file or directory is currently compressed. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* smbd: split out dos_mode debug print functionDavid Disseldorp2013-11-221-12/+31
| | | | | Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* vfs_btrfs: advertise per-file compression capabilityDavid Disseldorp2013-11-222-0/+20
| | | | | | | | | | Plumb into the get_fs_capabilities code path to advertise the FILE_FILE_COMPRESSION attribute. Also, document file compression support in the VFS btrfs man page. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* vfs_btrfs: fix copy-chunk dest unlock argsDavid Disseldorp2013-11-221-1/+1
| | | | | | | src_fsp used instead of dest_fsp for unlock. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* vfs_btrfs: add [GET/SET]_COMPRESSION handlersDavid Disseldorp2013-11-222-3/+118
| | | | | | | | | Translate such requests into FS_IOC_GETFLAGS and FS_IOC_SETFLAGS ioctls respectively. The module makes no distinction between compression types, only whether or not compression is enabled. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* smb2/ioctl: add support for FSCTL_[GET/SET]_COMPRESSIONDavid Disseldorp2013-11-221-2/+115
| | | | | | | | | | | In line with MS-FSCC 2.3.47, the FSCTL_SET_COMPRESSION ioctl allows remote SMB2 clients to enable and disable compression on a per-file or per-directory basis. Compression state can be retrieved using the FSCTL_GET_COMPRESSION request. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* vfs: add [GET/SET]_COMPRESSION hooksDavid Disseldorp2013-11-229-1/+205
| | | | | | | | | | | | | | | The VFS interfaces are sychronous, as the operations only modify meta-data. These hooks are dependent on support for transparent compression by the underlying filesystem - vfs_default returns INVALID_DEVICE_REQUEST. Support for other filesystems providing transparent comression, such as Btrfs and ZFS, can be added in future. The get_compression function takes fsp and smb_fname arguments. The smb_fname argument is needed due to the current dosmode() code-path. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* torture: attempt to set compression via SetInfoDavid Disseldorp2013-11-221-2/+101
| | | | | | | | Doing so should give a successful SetInfo response, however the attribute should not be set in subsequent GetInfo / ioctl responses. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* torture: test dir non-inherit in compress_inherit_disableDavid Disseldorp2013-11-221-8/+35
| | | | | | | | | | | Also add a note reguarding Windows 2008GM copy-chunk note: Windows 2008GM (non-R2) fails the copy-chunk src=dest overlap test, as it appears to use a different chunk copy algorithm to 208R2. Takeaway advice is to avoid using copy-chunk when the source and destination ranges overlap in the same file. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* selftest/s3: run smb2.ioctl against FS specific shareDavid Disseldorp2013-11-221-0/+4
| | | | | | | | Run the smb2.ioctl tests against the filesystem specific share. This adds ~12s to the selftest runtime on my SSD+Btrfs setup. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
* selftest/s3: expose share with FS applicable configDavid Disseldorp2013-11-221-1/+25
| | | | | | | | | | Attempt to determine the underlying test share filesystem using statfs. If the filesystem is detected and an applicable configuration is available, then apply it to a new filesystem specific share. Btrfs is currently the only VFS module to utilize this logic. Signed-off-by: David Disseldorp <ddiss@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
generated by cgit (git 2.34.1) at 2025-09-27 03:49:26 +0000