blobfs: start a poller to do cache buffer reclaim regularly

Currently we will free the cache buffers in the file R/W
data path, however we can start a thread to do the cache
buffer reclaim regulary,

blobfs: start a poller to do cache buffer reclaim regularly

Currently we will free the cache buffers in the file R/W
data path, however we can start a thread to do the cache
buffer reclaim regulary, so that we don't need to free
the cache buffer in the Rocksdb thread context. By
doing this, we can also remove the global cache buffer
lock in following patches.

Change-Id: Icb0fbc49baecd1e9d86a5fcfe400758dfc3c53a2
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/941
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>

show more ...

blobfs: try to free a valid file cache buffer when allocating new caches

When allocating a new cache buffer, we will pick up a file and try to
free cache buffers with this file, but sometimes the fi

blobfs: try to free a valid file cache buffer when allocating new caches

When allocating a new cache buffer, we will pick up a file and try to
free cache buffers with this file, but sometimes the file lock maybe
hold by other thread, so we use pthread_spin_trylock() here, when
pthread_spin_trylock() return a negative value, we will break the
loop, this is not efficient as the orginal logic, which may result
cache buffer allocation failure, so we will break the loop only when
there is a valid cache free.

Change-Id: I7be2fddb33c38dba466c2fcc2584ccf7ac16ea7a
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/950
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>

show more ...

blobfs: make cache_free_buffers() used for file deletion and blobfs unload

When there is no free cache buffers, we need to free some cache buffers
based on very simple algorithm for now, when deleti

blobfs: make cache_free_buffers() used for file deletion and blobfs unload

When there is no free cache buffers, we need to free some cache buffers
based on very simple algorithm for now, when deleting a file or unload
the blobfs, there must no dirty caches, there is no cache buffers after
calling spdk_tree_free_buffers(), so we will create another cache free
buffers function which only used in file read/write path.

When testing blobfs with db_bench, there maybe multiple threads will
call alloc_cache_memory_buffer() at same time, and spdk_fs_delete_file_async()
maybe called at same time, and it may release file->tree and file structure
without holding g_caches_lock, so here we will change the
alloc_cache_memory_buffer() to hold the g_caches_lock while releasing
the file cache buffers.

Fix issue #1133.

Change-Id: I115ff97113915630f539c3458c57116d9ff16179
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/668
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>

show more ...

blobfs: remove _file_read() function

_file_read() is only be called in spdk_file_read() API, and we only
need to count the children IOs sent to the backend, and skip the IOs
which hit the cache buff

blobfs: remove _file_read() function

_file_read() is only be called in spdk_file_read() API, and we only
need to count the children IOs sent to the backend, and skip the IOs
which hit the cache buffer.

Change-Id: I36e4a464e5810beaf30f1b580dc15ea3d3a68582
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/667
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

blobfs: write IO directly if it is lba aligned

UT also added for the block aligned write and verification
of the written data.

Change-Id: I8751f437c71dd7f3a4556fee420460d6dffdd4e5
Signed-off-by: Ga

blobfs: write IO directly if it is lba aligned

UT also added for the block aligned write and verification
of the written data.

Change-Id: I8751f437c71dd7f3a4556fee420460d6dffdd4e5
Signed-off-by: GangCao <gang.cao@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/478424
Community-CI: SPDK CI Jenkins <sys_sgci@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>

show more ...

blobfs: add result for set_cache_size

Since g_fs_cache_size only takes effect when creating
g_cache_pool, so spdk_fs_set_cache_size is only
permitted when cache pool is already freed or hasn't
been

blobfs: add result for set_cache_size

Since g_fs_cache_size only takes effect when creating
g_cache_pool, so spdk_fs_set_cache_size is only
permitted when cache pool is already freed or hasn't
been initialized.
Add a return value to indicate the result
of spdk_fs_set_cache_size.

Change-Id: I3828b136976d6f03f0751b2f20f68cd47c36ec04
Signed-off-by: Xiaodong Liu <xiaodong.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/471869
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

blobfs: use __wake_caller() to wake up the caller

__fs_create_file_done() is just used to wake up the caller
in synchronous API.

Change-Id: I08611eab74ab353021d5e9d7a33fe825f4c5a8c2
Signed-off-by:

blobfs: use __wake_caller() to wake up the caller

__fs_create_file_done() is just used to wake up the caller
in synchronous API.

Change-Id: I08611eab74ab353021d5e9d7a33fe825f4c5a8c2
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/464362
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>

show more ...

blobfs: return error to caller when loading a bs which isn't blobfs

Also release the IO channels with filesystem data structure.

Change-Id: I75a5d427fe4837ad7d85444d09ded9ab41e0ca8e
Signed-off-by:

blobfs: return error to caller when loading a bs which isn't blobfs

Also release the IO channels with filesystem data structure.

Change-Id: I75a5d427fe4837ad7d85444d09ded9ab41e0ca8e
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/464851
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>

show more ...

blobfs: use macro instead of hardcoded BLOBFS string

We used "BLOBFS" as the signature when loading from blobstore, so define
a macro here.

Change-Id: Iff504a36686c2bc08c90c49b8a505b5533b62508
Sign

blobfs: use macro instead of hardcoded BLOBFS string

We used "BLOBFS" as the signature when loading from blobstore, so define
a macro here.

Change-Id: Iff504a36686c2bc08c90c49b8a505b5533b62508
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/464850
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

blobfs: free sync request if got error with flush request allocation

Change-Id: I13bd02b1083816781980bcd7af7a6dd27235ed7d
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://

blobfs: free sync request if got error with flush request allocation

Change-Id: I13bd02b1083816781980bcd7af7a6dd27235ed7d
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/464360
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>

show more ...

blobfs: make internal asynchronous APIs as public APIs

SPDK blobfs has asynchronous APIs defined in blobfs_internal.h file,
as users may want to use them, so we remove them to the public .h file.

C

blobfs: make internal asynchronous APIs as public APIs

SPDK blobfs has asynchronous APIs defined in blobfs_internal.h file,
as users may want to use them, so we remove them to the public .h file.

Change-Id: I1835d97060101f6315a73cb8638b15ff7e13ba54
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/457547
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>

show more ...

blobfs: add writev/readv asynchronous APIs support

Change-Id: Id1172f546852fcf25c6d13cb63f9d875b02e768c
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.i

blobfs: add writev/readv asynchronous APIs support

Change-Id: Id1172f546852fcf25c6d13cb63f9d875b02e768c
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453493
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>

show more ...

blobfs: fix the semphore usage issue in spdk_file_read

We need to move sub_reads++ statement earlier. Otherwise if
the read fails, the sem_wait(&channel->sem) call number
is not correct.

Change-Id:

blobfs: fix the semphore usage issue in spdk_file_read

We need to move sub_reads++ statement earlier. Otherwise if
the read fails, the sem_wait(&channel->sem) call number
is not correct.

Change-Id: I05281a74bef78e4f80ce8d202b47849d1e6c2009
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453788
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>

show more ...

blobfs: Add trace for file metadata operation.

The purpose of is patch is to track whether there is
req allocation failure for those file metadata related
operation in order to find the bug.

Signed

blobfs: Add trace for file metadata operation.

The purpose of is patch is to track whether there is
req allocation failure for those file metadata related
operation in order to find the bug.

Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: I171a5b7fca214675fb1356ca650a01eeb20b4c8c
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/454829
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>

show more ...

blobfs: don't flush a partial buffer with no sync pending

We flush a cache buffer once it's filled. When the write
for that cache buffer has completed, we look to see if
there's more data to flush.

blobfs: don't flush a partial buffer with no sync pending

We flush a cache buffer once it's filled. When the write
for that cache buffer has completed, we look to see if
there's more data to flush. Currently if there's *any*
more data to flush, we will flush it, even if it's not
a full buffer.

That can hurt performance though. Ideally we only want
to flush a partial buffer if there's been an explicit
sync operation that requires that partial buffer to be
flushed. Otherwise we will end up writing the partial
buffer to disk, and then come back and write that data
again later when the buffer is full.

Add a new unit test to test for this condition. This
patch breaks one of the existing unit tests which was
designed specifically around a RocksDB failure condition.
Change that file_length unit test to now write exactly
one CACHE_BUFFER, which still tests the general logic
making sure that we don't confuse the amount of data
flushed with the value written to the file's length
xattr.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I83795fb45afe854b38648d0e0c1a7928219307a2

Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455698
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>

show more ...

blobfs: ensure length xattr is written even if all data is flushed

Data can get implicitly flushed as cache buffers are filled. But
the length xattr is only written in response to a sync or close
o

blobfs: ensure length xattr is written even if all data is flushed

Data can get implicitly flushed as cache buffers are filled. But
the length xattr is only written in response to a sync or close
operation. So we cannot just look at the amount of data flushed,
and ignore the sync operation if all of the data written has been
flushed - we still need to write the length xattr.

This also adds a unit test which reproduces the original problem.

Fixes issue #297.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Icca6ef4d1544f72e9bc31c4ee77d26b4b7f0cce4

Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455692
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>

show more ...

blobfs: removed unaligned memory access

Replaced (uint64_t *) dereference of (char *) variable with memcpy to
avoid unaligned memory access. It fixes the following error reported by
ubsan:

spdk/lib

blobfs: removed unaligned memory access

Replaced (uint64_t *) dereference of (char *) variable with memcpy to
avoid unaligned memory access. It fixes the following error reported by
ubsan:

spdk/lib/blobfs/blobfs.c:1492:2: runtime error: load of misaligned
address 0x00000047a136 for type 'uint64_t', which requires 8 byte
alignment

Change-Id: I5f3f28a52b8cbe3eb30394ae7ebd0e64b683db57
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/454886
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>

show more ...

blobfs: Add the trace support for spdk_file_delete.

Change-Id: Ida9d3e6e6b7a35cd924726294feff934a1246a6d
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spd

blobfs: Add the trace support for spdk_file_delete.

Change-Id: Ida9d3e6e6b7a35cd924726294feff934a1246a6d
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/454212
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Xiaodong Liu <xiaodong.liu@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>

show more ...

blobfs: pass sync_req instead of file as the parameter

This avoids to get the first sync_req in the list, instead of
make sure which sync_req we want to make it complete.

Also release the request e

blobfs: pass sync_req instead of file as the parameter

This avoids to get the first sync_req in the list, instead of
make sure which sync_req we want to make it complete.

Also release the request early.

Change-Id: If6f9b2ab10f0a93e6c45088d33f2aa78d25f66cb
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/452136
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>

show more ...

blobfs: fix the wrong parameter in spdk_file_close

According to my code understanding, should be assigned
itself instead of req.

Change-Id: I0af4d56f8029269e0632bb6be6c9d912c7106cf8
Signed-off-by:

blobfs: fix the wrong parameter in spdk_file_close

According to my code understanding, should be assigned
itself instead of req.

Change-Id: I0af4d56f8029269e0632bb6be6c9d912c7106cf8
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/452120
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>

show more ...

blobfs: remove trace short names

They were removed from the trace API and BlobFS doesn't
compile right now.

Change-Id: Ie98339f19353e4db901af551c706f57c18b89fc5
Signed-off-by: Darek Stojaczyk <dari

blobfs: remove trace short names

They were removed from the trace API and BlobFS doesn't
compile right now.

Change-Id: Ie98339f19353e4db901af551c706f57c18b89fc5
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/452863
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

blobfs: add trace events

For now, just do events for open, close and start/end
for writing the length xattr. These will be useful for
debugging the intermittent RocksDB failures.

Note these events

blobfs: add trace events

For now, just do events for open, close and start/end
for writing the length xattr. These will be useful for
debugging the intermittent RocksDB failures.

Note these events may only be recorded on the polling
SPDK thread. SPDK events cannot be recorded on
non-SPDK threads (i.e. RocksDB user or background threads).

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I72fb302db23b399f583f56194c237340587f3027

Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/452262
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>

show more ...

blobfs: track num outstanding requests per channel

This will be used in next patch to ensure synchronous
threads wait until any background operations are done
before freeing its request memory, and

blobfs: track num outstanding requests per channel

This will be used in next patch to ensure synchronous
threads wait until any background operations are done
before freeing its request memory, and also to print
errors if internal channels have outstanding
requests when freed.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I24bc8096f84b70270f87b051f9034ed08e639500

Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/451780
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

blobfs: use wrapper function to free fs resources

Change-Id: Ia9c430c512e8c9a3195e2bfb28adb0526174070a
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io

blobfs: use wrapper function to free fs resources

Change-Id: Ia9c430c512e8c9a3195e2bfb28adb0526174070a
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/450982
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

blobfs: fix up the wrong usage of delete API

blob_delete_cb() takes spdk_fs_request as the input
parameter.

Change-Id: Ie6150e7d31d187296a448e82784e2ac2fecfe52c
Signed-off-by: Changpeng Liu <changp

blobfs: fix up the wrong usage of delete API

blob_delete_cb() takes spdk_fs_request as the input
parameter.

Change-Id: Ie6150e7d31d187296a448e82784e2ac2fecfe52c
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/450895
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>

show more ...