[XS] Change in ...nfs-ganesha[next]: FSAL_CEPH: Avoid duplicate refcounting of cmount
by Sachin Punadikar (GerritHub)
Sachin Punadikar has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1193055?usp=email )
Change subject: FSAL_CEPH: Avoid duplicate refcounting of cmount
......................................................................
FSAL_CEPH: Avoid duplicate refcounting of cmount
Current code initialises cmount refcount to 1 and then it increases
again in later code flow. Also when same cmount path is being used, then
refcounting was done twice. Corrected the code.
Change-Id: I23456ac94a64b332ad563be5f91f637d330911cf
Signed-off-by: Sachin Punadikar <sachin.punadikar(a)ibm.com>
---
M src/FSAL/FSAL_CEPH/main.c
1 file changed, 1 insertion(+), 2 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/55/1193055/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1193055?usp=email
To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings?usp=email
Gerrit-MessageType: newchange
Gerrit-Project: ffilz/nfs-ganesha
Gerrit-Branch: next
Gerrit-Change-Id: I23456ac94a64b332ad563be5f91f637d330911cf
Gerrit-Change-Number: 1193055
Gerrit-PatchSet: 1
Gerrit-Owner: Sachin Punadikar <sachin.punadikar(a)ibm.com>
1 month
Announce Push of V6-dev.9
by Frank Filz
Branch next
Tag:V6-dev.9
I did a mid-week merge to move Dipit's ID Mapper patch set along.
Merge Highlights
* This merge is a major rework of the idmapping caching
Signed-off-by: Frank S. Filz <ffilzlnx(a)mindspring.com>
Contents:
2cf6d40fa Frank S. Filz V6-dev.9
5aca72270 Dipit Grover Idmapper: Refactor negative_cache_entry to add an
explicit name-buffer
89833338d Dipit Grover Idmapper: Add function to reap expired user-groups
cached entries
38b1641bb Dipit Grover Idmapper: Limit max user-groups cache entries, and
add a fifo queue
f0cccdd3b Dipit Grover Idmapper: Add reaper function for negative cache
e67c1b835 Dipit Grover Idmapper: Introduce reaper to remove expired user,
group cache entries
6036cdd60 Dipit Grover Idmapper: Integrate group negative-cache during name
resolution
9a8bf8b3d Dipit Grover Idmapper: Introduce negative cache for group entries
4b3fe40fc Dipit Grover Idmapper: Integrate user negative-cache during
name/principal resolution
f2fb94689 Dipit Grover Idmapper: Introduce negative cache for user entries
f86dc42c1 Dipit Grover Idmapper: Remove redundant else block from name2id
function
c7364bb86 Dipit Grover Idmapper: Limit max group-cache entries, and add a
fifo queue
1f2c676f1 Dipit Grover Idmapper: Limit max user-cache entries, and add a
fifo queue
a742430d8 Dipit Grover Idmapper: Extract user and group cache removal into
separate functions
0aac4af74 Dipit Grover Idmapper: Also clear user-groups cache when
performing cleanup
8466a7248 Dipit Grover Idmapper: Add idmapped-group's time validity to
DIRECTORY_SERVICES config
0ef408657 Dipit Grover Idmapper: Add idmapped-user's time validity to
DIRECTORY_SERVICES config
0d0109a76 Dipit Grover Idmapper: Move gsh-buffer comparator function to
common_utils
2c38ded8b Dipit Grover Idmapping: Move DomainName from NFSv4 to new
DIRECTORY_SERVICES config section
1 month
Large patch sets
by Frank Filz
Hello all,
Recently we have been getting a number of large patch sets. Please note that
a large patch set often makes it harder to review. It also overwhelms the CI
which submits multiple jobs for each patch submitted (and updated). I'm
sorry if I'm not getting to review all the patches, there are a lot. And my
GerritHub dashboard is getting unwieldy.
If I could make some suggestions on consolidating patches:
If you are introducing a bunch of new functions or config parameters and
have patches that provide those, but utilize them in a separate patch,
consider if you can consolidate all the new stuff into a single patch.
In general, try and consolidate changes, especially small ones. A major
change, even if other changes are related, may well make sense to be in a
separate patch.
It is nice when a patch doesn't make a second change to lines changed in a
previous patch.
If you are making a FSAL API change that is small, consider one patch to
update all the FSALs. On the other hand, for a larger change, one patch per
FSAL makes it easier to assure that each FSAL's stake holders have reviewed
the change. In this case, a larger patch set is understandable.
If you have patches that fix bugs or other stuff that isn't exactly
pertinent to your changes, it can be helpful to have those patches at the
front of your patch set (farthest from HEAD). I have sometimes merged these
while still discussing the rest of the patch set.
On another note, I am seeing a lot of checkpatch failures, please make sure
you have the checkpatch commit hook installed as per CONTRIBUTING_HOWTO.txt.
One of the patch sets submitted is a suggestion to change to clang, which I
am in favor of, and we will make that change when the dust settles some.
Once we make that change, the commit hook will change, but until then,
please conform to checkpatch.
Thanks
Frank
1 month, 1 week
Announce Push of V6-dev.8
by Frank Filz
Branch next
Tag:V6-dev.8
Merge Highlights
* Reduce the default value of Readdir_Res_Size to 32KB
* two delegation fixes
* four fd_lru fixes
* Decrement Client Ref, if reserve lease or expire fails
* FSAL_CEPH : Add warnings for export creation
* GPFS: FSAL's state_free function called by free_state doesn't actually
free
* GTest - fix tests for new FSAL APIs
* Add PTHREAD_COND_broadcast macro
* fsal_fd fd_work/io_work needs separate condition variables
Signed-off-by: Frank S. Filz <ffilzlnx(a)mindspring.com>
Contents:
40483a71e Frank S. Filz V6-dev.8
f74d92ee5 Daniel Gryniewicz GTest - fix tests for new FSAL APIs
c03f0d368 Yogendra Charya Tangalapally GPFS: FSAL's state_free function
called by free_state doesn't actually free
c84e8ff17 Martin Schwenke doc: Update default for Readdir_Res_Size to 32KB
05773ad01 Sachin Punadikar FSAL_CEPH : Add warnings for export creation
45eccd509 Frank S. Filz fsal_fd fd_work/io_work needs separate condition
variables
38631af28 Frank S. Filz Add PTHREAD_COND_broadcast macro
8c21c4f8f Rojin George Free delegation recall context after cleaning expired
clients
9401f7b9e Rojin George Decrement Client Ref, if reserve lease or expire
fails
08f294a40 Rojin George Handle parallel delegation recalls for the same
fsal_objs
aaad4fe20 Rojin George Insert global FDs of reopened files into fd lru
d25691223 Rojin George Provide FSALs option to remove global FDs from fd_lru
upon failures
9e74145f3 Rojin George Expose bump_fd_lru to FSALs to bump up if reusing
global FDs
7be4ad100 Rojin George Insert to fd_lru only if open/reopen succeeds
1 month, 1 week
pthread_rwlock on el8
by gaurav gangalwar
Hello,
Need some thoughts on the use of pthread_rwlock in Ganesha with zfs as
backend filesystem on el8
We see regression with specfs swbuild workload on el8 with same Ganesha
code that was performing better on el7.
To triage the issue we captured perf stats and using flamegraph found that
Ganesha is spending a lot of cpu cycles in pthread_rwlock on el8 as
compared to el7.
If we convert the pthread_rwlock that shows up in perf stats to
pthread_mutex, we no longer see this issue.
Locks that we converted to pthread_mutex are fsm_lock, attr_lock, obj_lock,
cih_lock and mdc_exp_lock
This issue gets more severe as the number of threads grows in Ganesha, more
threads, higher CPU utilization and impact on IOPs.
Has anyone faced this issue on el8?
This is patch for reference that is working for us
https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1191904
Regards,
Gaurav Gangalwar
1 month, 1 week
[L] Change in ...nfs-ganesha[next]: To fix issue with pthread_rwlock on el8, converted some of pthread_rw...
by Gaurav (GerritHub)
Gaurav has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1191904?usp=email )
Change subject: To fix issue with pthread_rwlock on el8, converted some of pthread_rwlock to pthread_mutex.
......................................................................
To fix issue with pthread_rwlock on el8, converted some of pthread_rwlock
to pthread_mutex.
Change-Id: I38bd5ba24a3f7ba6f301cf6597102a7a31e41210
Signed-off-by: Gaurav Gangalwar <gaurav.gangalwar(a)gmail.com>
---
M src/FSAL/FSAL_CEPH/handle.c
M src/FSAL/FSAL_CEPH/mds.c
M src/FSAL/FSAL_GLUSTER/handle.c
M src/FSAL/FSAL_GPFS/file.c
M src/FSAL/FSAL_KVSFS/kvsfs_file.c
M src/FSAL/FSAL_LIZARDFS/handle.c
M src/FSAL/FSAL_MEM/mem_handle.c
M src/FSAL/FSAL_PSEUDO/handle.c
M src/FSAL/FSAL_RGW/handle.c
M src/FSAL/FSAL_VFS/file.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_export.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_file.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_handle.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_hash.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_hash.h
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_int.h
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_main.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_up.c
M src/FSAL/commonlib.c
M src/FSAL/default_methods.c
M src/FSAL/fsal_manager.c
M src/include/FSAL/fsal_commonlib.h
M src/include/fsal_api.h
25 files changed, 154 insertions(+), 154 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/04/1191904/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1191904?usp=email
To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings?usp=email
Gerrit-MessageType: newchange
Gerrit-Project: ffilz/nfs-ganesha
Gerrit-Branch: next
Gerrit-Change-Id: I38bd5ba24a3f7ba6f301cf6597102a7a31e41210
Gerrit-Change-Number: 1191904
Gerrit-PatchSet: 1
Gerrit-Owner: Gaurav <gaurav.gangalwar(a)gmail.com>
1 month, 1 week
