[XS] Change in ...nfs-ganesha[next]: Unblock important log entries
by Sachin Punadikar (GerritHub)
Sachin Punadikar has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1197796?usp=email )
Change subject: Unblock important log entries
......................................................................
Unblock important log entries
By default FATAL and WARN messages are getting logged. At certain places
these messages are blocked by checking FULL_DEBUG enabled for the
related component. This change unblockes 2 such important messages.
Change-Id: I9cf17178ec268fd4a3455aa02683881882e19cfc
Signed-off-by: Sachin Punadikar <sachin.punadikar(a)ibm.com>
---
M src/SAL/nfs4_clientid.c
1 file changed, 2 insertions(+), 2 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/96/1197796/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1197796?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: I9cf17178ec268fd4a3455aa02683881882e19cfc
Gerrit-Change-Number: 1197796
Gerrit-PatchSet: 1
Gerrit-Owner: Sachin Punadikar <sachin.punadikar(a)ibm.com>
5 months, 2 weeks
[S] Change in ...nfs-ganesha[next]: Add validation to Server stats to get latest info
by Rojin George (GerritHub)
Rojin George has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1197444?usp=email )
Change subject: Add validation to Server stats to get latest info
......................................................................
Add validation to Server stats to get latest info
To list the client with active connections per protocol, changing
the logics to return protocol availability by making changes
based on the counters. The user/admin who wants to get such fresh
updated view could run ResetStats DBus prior to ShowClients .
Change-Id: I2014c86a490d46e685a6fc06eb4ad43414f43429
Signed-off-by: Rojin George <itsmerojin(a)gmail.com>
---
M src/support/server_stats.c
1 file changed, 10 insertions(+), 8 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/44/1197444/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1197444?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: I2014c86a490d46e685a6fc06eb4ad43414f43429
Gerrit-Change-Number: 1197444
Gerrit-PatchSet: 1
Gerrit-Owner: Rojin George <itsmerojin(a)gmail.com>
5 months, 2 weeks
SOLVED! was RE: PLEASE READ - NO MERGE THIS WEEK - NEED TO PULL BACK AND ANALYZE FAILURES
by Frank Filz
I solved which patch was the culprit and have just pushed V6-dev.16.
Thanks for your patience and code contributions.
Frank Filz
From: Frank Filz [mailto:ffilzlnx@mindspring.com]
Sent: Friday, July 5, 2024 5:34 PM
To: devel(a)lists.nfs-ganesha.org
Subject: [NFS-Ganesha-Devel] PLEASE READ - NO MERGE THIS WEEK - NEED TO PULL
BACK AND ANALYZE FAILURES
Sorry, I just spent several hours trying to get a successful merge. I keep
getting Ganesha crashing during the pynfs 4.0 RPLY8 test case. I thought I
had it narrowed down to a particular patch, so rebuilt V6-dev.16 without
that patch, and started a full pynfs 4.0 test run which promptly caused
Ganesha to crash while running RPLY8.
Based on the test case I actually suspect the courteous server support is
causing the crash (it's a test that would be impacted by courteous server -
establishing a lock then not renewing least for 2 lease periods).
We are going to have step back and deeply analyze code to determine the
cause of these crashes.
In the meantime, I am rescinding all +2 code reviews. I will leave +1 code
reviews in place.
Thanks
Frank Filz
5 months, 2 weeks
Announce Push of V6-dev.16
by Frank Filz
Branch next
Tag:V6-dev.16
Merge Highlights
* In nfs4_Check_Stateid don't get obj2 reference if no fsal_obj
* Expose some functions for unit tests
* Support free disk space reporting for project quotas
* src/include/log.h: LogFatal calls abort.
* src/monitoring/prometheus-cpp-lite: Added new submodule with better
performance.
* Use prometheus-cpp-lite and extract to a separate lib.
* Convert some existing metrics to static and introduce new static metrics
* nfs_core: Make nfs_get_evchannel_id public
* hashtable: Add more hashtable needed functions, to allow FSAL to create
hash table
* nfs_rpc_callback: Assert in case no completion function is set as we will
have a cb resource leak
* nfs4_op_close: Release locks on close
* sal_data.h: Factor out lock protocol type from blocking state
* state_lock.c: Implement nfsv4 blocking locks support
Signed-off-by: Frank S. Filz <ffilzlnx(a)mindspring.com>
Contents:
627daf37b Frank S. Filz V6-dev.16
d618b7a5a Ofir Vaisnhtein state_lock.c: Implement nfsv4 blocking locks
support
d857fc862 Ofir Vaisnhtein sal_data.h: Factor out lock protocol type from
blocking state
4fafc1af7 Ofir Vaisnhtein nfs4_op_close: Release locks on close
12a9abd7f Ofir Vainshtein nfs_rpc_callback: Assert in case no completion
function is set as we will have a cb resource leak
dc35e38fe Ofir Vainshtein hashtable: Add more hashtable needed functions, to
allow FSAL to create hash table
1e4571814 Ofir Vainshtein nfs_core: Make nfs_get_evchannel_id public
09b660eb4 Yoni Couriel RPCAL/connection_manager_metrics.c: Added connection
manager static metrics.
b7f305286 Yoni Couriel idmapper/idmapper_monitoring.c: Added static metrics
to idmapper.
681fe5b79 Yoni Couriel MainNFSD/nfs_metrics.c: Added NFSv4 static metrics.
11b564c00 Yoni Couriel monitoring/monitoring.cc: Added
Enable_Dynamic_Metrics param to disable dynamic metrics.
b97d5058e Yoni Couriel MainNFSD/nfs_metrics.c: Convert RPC metrics from
dynamic to static.
ab9541451 Yoni Couriel monitoring/monitoring.cc: Introduce new STATIC
metrics API.
514ed8004 Yoni Couriel Use prometheus-cpp-lite and extract to a separate
lib.
4f96c6a8e Yoni Couriel src/monitoring/prometheus-cpp-lite: Added new
submodule with better performance.
130ecc6e4 Yoni Couriel src/include/log.h: LogFatal calls abort.
1ae609bb0 Yoni Couriel .gitignore: Added .vscode
eafca8918 Trupti Shete Support free disk space reporting for project quotas
c81a2d13b David Rieber Expose initialize_fsal_lock and destroy_fsal_lock
functions. We use these in unit tests.
e8c675424 David Rieber Expose a nfs_prereq_init_mutexes function to be able
to initialize mutexes during unit tests.
95d0a9b27 Frank S. Filz In nfs4_Check_Stateid don't get obj2 reference if no
fsal_obj
5 months, 2 weeks
PLEASE READ - NO MERGE THIS WEEK - NEED TO PULL BACK AND ANALYZE FAILURES
by Frank Filz
Sorry, I just spent several hours trying to get a successful merge. I keep
getting Ganesha crashing during the pynfs 4.0 RPLY8 test case. I thought I
had it narrowed down to a particular patch, so rebuilt V6-dev.16 without
that patch, and started a full pynfs 4.0 test run which promptly caused
Ganesha to crash while running RPLY8.
Based on the test case I actually suspect the courteous server support is
causing the crash (it's a test that would be impacted by courteous server -
establishing a lock then not renewing least for 2 lease periods).
We are going to have step back and deeply analyze code to determine the
cause of these crashes.
In the meantime, I am rescinding all +2 code reviews. I will leave +1 code
reviews in place.
Thanks
Frank Filz
5 months, 3 weeks
[S] Change in ...nfs-ganesha[next]: In nfs4_Check_Stateid don't get obj2 reference if no fsal_obj
by Frank Filz (GerritHub)
Frank Filz has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1197070?usp=email )
Change subject: In nfs4_Check_Stateid don't get obj2 reference if no fsal_obj
......................................................................
In nfs4_Check_Stateid don't get obj2 reference if no fsal_obj
In the call to get_state_obj_export_owner_refs, don't ask for obj2 if
fsal_obj is NULL. We don't need the obj2 reference, and we may not have
an op_context to be able to release the reference.
This is indeed the case when FREE_STATEID is called without having done
a PUTFH. Note that subsequently, FREE_STATEID DOES make it's own call
to get_state_obj_export_owner_refs, but it uses the export returned to
set up an op_context thus allowing for the put_ref.
Fixes github issue #1151
Change-Id: I71bca94c03c120265d610d43784ca879950f18bc
Signed-off-by: Frank S. Filz <ffilzlnx(a)mindspring.com>
---
M src/SAL/nfs4_state_id.c
1 file changed, 13 insertions(+), 7 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/70/1197070/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1197070?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: I71bca94c03c120265d610d43784ca879950f18bc
Gerrit-Change-Number: 1197070
Gerrit-PatchSet: 1
Gerrit-Owner: Frank Filz <ffilzlnx(a)mindspring.com>
5 months, 3 weeks