Hi all,
I am quit new to this project, so I do not really know how to debug. I
simply started
corosync with:
# gdb corosync
> run -f
The I start the clvm-openais daemon, which uses the lock service. The
service aquires a lock, and then calls:
saLckResourceClose(linfo->res_handle);
everything l
This is used nowhere in the code.
Fabio
Index: logsys.c
===
--- logsys.c (revision 1989)
+++ logsys.c (working copy)
@@ -161,12 +161,6 @@
#define FDTAIL_INDEX (flt_data_size + 1)
-struct log_data {
- unsigned int syslog_pos;
- u
--- openais-0.94/test/Makefile.org 2009-04-01 10:31:55.0
-0400
+++ openais-0.94/test/Makefile 2009-04-01 10:33:00.0 -0400
@@ -55,7 +55,7 @@
testckpt: testckpt.o sa_error.o
$(CC) $(LDFLAGS) -o testckpt testckpt.o $(LIBS)
-ckptbench: ckptbench.o
+ckptbench: ckptbench.
> Try to create this directory before running corosync and see if that
> fixes your problem.
>
> Please report back if that is the root cause, and I'll work up a
better
> notification patch.
Wow, that was a quick solution :-) Here is the output from strace:
mkdir("/usr/var/lib/corosync", 0700)
Dietmar,
Current trunk builds/runs for me.
Perhaps you have selinux enabled on your system? There may be a problem
where corosync tries to create the /var/lib/corosync directory and is
unable to because selinux doesn't allow it. Unfortunately it appears no
proper warning is given to the user in
Steven Dake wrote:
> looks good
committed.
___
Openais mailing list
Openais@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/openais
looks good
regards
-steve
On Wed, 2009-04-01 at 08:35 +0200, Fabio M. Di Nitto wrote:
> correct patch in attachment without the mainconfig.c changes that are
> part of another posted patch.
>
> Fabio
>
> On Wed, 2009-04-01 at 08:33 +0200, Fabio M. Di Nitto wrote:
> > doesn't matter to me.. new
this implementation looks good and the proper choice. I don't want to
change the upstream api since this is a callback.
Regards
-steve
On Wed, 2009-04-01 at 18:24 +0200, Jim Meyering wrote:
> Jim Meyering wrote:
> > Here's what happened when I started adjusting the APIs in confdb.h.
> > Changing
I already merged your previous patch. Can you rework the defines into a
new patch? Thanks!
Regards
-steve
On Wed, 2009-04-01 at 12:30 +0200, Jan Friesse wrote:
> Attached is patch with included suggested changes.
>
> Thanks Jim.
>
> Jim Meyering wrote:
> > Jan Friesse wrote:
> >> Attached pat
looks good
regards
-steve
On Wed, 2009-04-01 at 11:47 +0200, Jim Meyering wrote:
> There are still too many warnings ;-)
> I missed the fact that my recent changes introduced one new warning.
> Here's the fix:
>
> From fa2983e1e57fd4d772b76635e0625ed8c294959d Mon Sep 17 00:00:00 2001
> From: Jim
This patchset looks good but we are holding API changes until after we
release a bugfix version of corosync/oa for the release of cluster-3.
Please keep this handy for a weekish until that blockage is stopped.
Regards
-steve
On Wed, 2009-04-01 at 10:00 +0200, Jim Meyering wrote:
> Three patches
thanks for the patch!
merging
regards
-steve
On Wed, 2009-04-01 at 11:05 +0200, Jan Friesse wrote:
> Attached patch solves problem with running corosync as ais user.
>
> Main problem was hidden in reading aisexec section. If this section
> exists in corosync.conf, everything works, but in other
Jan Friesse wrote:
> Attached patch solves possible problems of overflow buffer, which can
> happened with sprintf. I tried to change only files where it makes sense
> (so basically, where in format is something like %s) and I hope I catch
> all cases.
>
>
> Index: test/sa_error.c
...
> -
Jim Meyering wrote:
> Here's what happened when I started adjusting the APIs in confdb.h.
> Changing void* to const void* propagated, of course.
> Changing int to size_t exposed the problem that
> tail_key_changed (and two other functions nearby)
> would mistakenly print non-printing (aka garbage)
Here's what happened when I started adjusting the APIs in confdb.h.
Changing void* to const void* propagated, of course.
Changing int to size_t exposed the problem that
tail_key_changed (and two other functions nearby)
would mistakenly print non-printing (aka garbage)
or incomplete object, key, val
>From 6386a7040c229961f3485bb9302f041ff0af0051 Mon Sep 17 00:00:00 2001
From: Jim Meyering
Date: Wed, 1 Apr 2009 18:29:24 +0200
Subject: [PATCH] confdb.h: continue API changes: const+size_t
* include/corosync/confdb.h (confdb_object_create_notify_fn_t):
(confdb_object_delete_notify_fn_t):
* tool
Hi all,
I sucessfully compiled corosync-0.94/openais-0.93, and even managed to
create a cluster
with 4 nodes - everything worked as expected.
But it does not work when I try corosync-0.95/openais-0.94 form svn,
using exactly the same configuration:
# # corosync-objctl
service.name=openais_clm
s
Jan Friesse wrote:
> I hope last fix of that patch.
Fine by me.
___
Openais mailing list
Openais@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/openais
Attached patch solves possible problems of overflow buffer, which can
happened with sprintf. I tried to change only files where it makes sense
(so basically, where in format is something like %s) and I hope I catch
all cases.
Index: test/sa_error.c
I hope last fix of that patch.
Jim Meyering wrote:
> Jan Friesse wrote:
>> Attached is patch with included suggested changes.
>
> Looks good.
>
> ...
>> +The default is ais.
>
> one more: s/is/for each is/
Index: exec/mainconfig.c
==
For the record:
15:19 < meyering> so consider your ACK'd and I'll try to stay away from
IRC
Committed revision 1987.
Fabio
On Wed, 2009-04-01 at 13:59 +0200, Fabio M. Di Nitto wrote:
> On Wed, 2009-04-01 at 13:31 +0200, Fabio M. Di Nitto wrote:
> > On Wed, 2009-04-01 at 13:15 +0200, Fabio M. Di
On Wed, 2009-04-01 at 13:31 +0200, Fabio M. Di Nitto wrote:
> On Wed, 2009-04-01 at 13:15 +0200, Fabio M. Di Nitto wrote:
> > On Wed, 2009-04-01 at 11:09 +0200, Fabio M. Di Nitto wrote:
> > > On Wed, 2009-04-01 at 10:40 +0200, Fabio M. Di Nitto wrote:
> > > > Hi,
> > > >
> > > > this patch replace
On Wed, 2009-04-01 at 13:15 +0200, Fabio M. Di Nitto wrote:
> On Wed, 2009-04-01 at 11:09 +0200, Fabio M. Di Nitto wrote:
> > On Wed, 2009-04-01 at 10:40 +0200, Fabio M. Di Nitto wrote:
> > > Hi,
> > >
> > > this patch replaces the one posted the 31st Of March and it's going to
> > > be my last ta
On 2009-03-19T15:27:30, Dominik Klein wrote:
> Hi
>
> I am using the latest whitetank code (1754 2009-03-13 20:47:05 +0100)
> with pacemaker on a pair of x86_64 opensuse 11.1 boxes and I am seeing
> an openais restart issue.
>
> When I use
> /etc/init.d/openais restart
> to restart the cluster,
On Wed, 2009-04-01 at 11:09 +0200, Fabio M. Di Nitto wrote:
> On Wed, 2009-04-01 at 10:40 +0200, Fabio M. Di Nitto wrote:
> > Hi,
> >
> > this patch replaces the one posted the 31st Of March and it's going to
> > be my last take on it. I think I finally found nirvana and imho good
> > enough to be
Jan Friesse wrote:
> Attached is patch with included suggested changes.
Looks good.
...
> +The default is ais.
one more: s/is/for each is/
___
Openais mailing list
Openais@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/
Attached is patch with included suggested changes.
Thanks Jim.
Jim Meyering wrote:
> Jan Friesse wrote:
>> Attached patch solves problem with running corosync as ais user.
>>
>> Main problem was hidden in reading aisexec section. If this section
>> exists in corosync.conf, everything works, but i
Jan Friesse wrote:
> Attached patch solves problem with running corosync as ais user.
>
> Main problem was hidden in reading aisexec section. If this section
> exists in corosync.conf, everything works, but in other cases,
> main_config->uid/gid are initialized to 0 (so only root:root) can run
> co
There are still too many warnings ;-)
I missed the fact that my recent changes introduced one new warning.
Here's the fix:
>From fa2983e1e57fd4d772b76635e0625ed8c294959d Mon Sep 17 00:00:00 2001
From: Jim Meyering
Date: Wed, 1 Apr 2009 11:46:43 +0200
Subject: [PATCH] corosync-objctl.c: remove a j
Jan Friesse wrote:
> Attached patch solves problem with running corosync as ais user.
>
> Main problem was hidden in reading aisexec section. If this section
> exists in corosync.conf, everything works, but in other cases,
> main_config->uid/gid are initialized to 0 (so only root:root) can run
> c
On Wed, 2009-04-01 at 10:40 +0200, Fabio M. Di Nitto wrote:
> Hi,
>
> this patch replaces the one posted the 31st Of March and it's going to
> be my last take on it. I think I finally found nirvana and imho good
> enough to be out there :)
>
> - add insert_into_buffer static function. This is a b
Attached patch solves problem with running corosync as ais user.
Main problem was hidden in reading aisexec section. If this section
exists in corosync.conf, everything works, but in other cases,
main_config->uid/gid are initialized to 0 (so only root:root) can run
corosync.
Another problem, what
On Wed, 2009-04-01 at 10:16 +0200, Jim Meyering wrote:
> From: Jim Meyering
>
> * tools/corosync-fplay.c (records_printed): Remove.
> ---
> tools/corosync-fplay.c |2 --
> 1 files changed, 0 insertions(+), 2 deletions(-)
>
> diff --git a/tools/corosync-fplay.c b/tools/corosync-fplay.c
> ind
On Wed, 2009-04-01 at 10:16 +0200, Jim Meyering wrote:
> From: Jim Meyering
>
> * tools/corosync-fplay.c: Make many parameters and local pointers const.
> ---
> tools/corosync-fplay.c | 158
>
> 1 files changed, 80 insertions(+), 78 deletions(-)
On Wed, 2009-04-01 at 10:16 +0200, Jim Meyering wrote:
> From: Jim Meyering
>
> * tools/corosync-fplay.c (sync_printer_nada): Remove function.
> * tools/corosync-objctl.c (get_child_name): Remove function.
> * test/evsverify.c (msg): Remove unused variable.
> ---
> test/evsverify.c|2
On Wed, 2009-04-01 at 10:16 +0200, Jim Meyering wrote:
> From: Jim Meyering
>
> * test/evsbench.c: Declare functions/vars static.
> * test/evsverify.c: Likewise.
> * test/testcpg2.c: Likewise.
> * test/testcpg.c: Likewise.
> * test/logsysbench.c: Likewise.
> * test/cpgbench.c: Likewise.
> * exec/
On Wed, 2009-04-01 at 10:16 +0200, Jim Meyering wrote:
> From: Jim Meyering
>
> * exec/main.c (sigintr_handler, serialize_lock, serialize_unlock):
> (serialize_lock, serialize_unlock):
> * exec/sync.c (sync_start_init, sync_callbacks_load):
> * exec/vsf_ykd.c (ykd_state_init):
> * lcr/uis.c (cmd1
On Wed, 2009-04-01 at 10:16 +0200, Jim Meyering wrote:
> From: Jim Meyering
>
> ---
> test/testevs.c | 20 ++--
> 1 files changed, 10 insertions(+), 10 deletions(-)
>
> diff --git a/test/testevs.c b/test/testevs.c
Good for me.
Fabio
_
Fabio M. Di Nitto wrote:
> correct patch in attachment without the mainconfig.c changes that are
> part of another posted patch.
...
This patch looks fine to me.
However, for the record, there are nasty bugs in the vicinity:
- nothing checks for buffer overflow via too-large output_buffer_idx
On Wed, 2009-04-01 at 10:16 +0200, Jim Meyering wrote:
> From: Jim Meyering
>
> * exec/timer.c: Include "timer.h".
> * exec/timer.h (corosync_timer_init): Make return type in prototype
> match the one in the function definition.
> ---
> exec/timer.c |1 +
> exec/timer.h |6 +++---
> 2 fi
On Wed, 2009-04-01 at 10:16 +0200, Jim Meyering wrote:
> From: Jim Meyering
>
> ---
> exec/apidef.c |1 +
> 1 files changed, 1 insertions(+), 0 deletions(-)
>
> diff --git a/exec/apidef.c b/exec/apidef.c
> index 951a6a1..7148042 100644
> --- a/exec/apidef.c
> +++ b/exec/apidef.c
> @@ -52,6
On Wed, 2009-04-01 at 10:16 +0200, Jim Meyering wrote:
> From: Jim Meyering
>
> * exec/objdb.c (objdb_wrlock, objdb_rdlock, objdb_rdunlock):
> (objdb_wrunlock):
> * services/cfg.c (send_shutdown, check_shutdown_status):
> * services/votequorum.c (send_expectedvotes_notification):
> * tools/corosy
Hi,
this patch replaces the one posted the 31st Of March and it's going to
be my last take on it. I think I finally found nirvana and imho good
enough to be out there :)
- add insert_into_buffer static function. This is a big helper to add
new logsys parameters anywhere in the current format_buff
From: Jim Meyering
* test/evsbench.c: Declare functions/vars static.
* test/evsverify.c: Likewise.
* test/testcpg2.c: Likewise.
* test/testcpg.c: Likewise.
* test/logsysbench.c: Likewise.
* test/cpgbench.c: Likewise.
* exec/sync.c: Likewise.
---
exec/sync.c|3 ++-
test/cpgbench.c
From: Jim Meyering
* tools/corosync-fplay.c: Make many parameters and local pointers const.
---
tools/corosync-fplay.c | 158
1 files changed, 80 insertions(+), 78 deletions(-)
diff --git a/tools/corosync-fplay.c b/tools/corosync-fplay.c
index c
From: Jim Meyering
---
test/testevs.c | 20 ++--
1 files changed, 10 insertions(+), 10 deletions(-)
diff --git a/test/testevs.c b/test/testevs.c
index 55e0d34..16b1e87 100644
--- a/test/testevs.c
+++ b/test/testevs.c
@@ -7,7 +7,7 @@
* Author: Steven Dake (sd...@redhat.com)
From: Jim Meyering
* exec/main.c (sigintr_handler, serialize_lock, serialize_unlock):
(serialize_lock, serialize_unlock):
* exec/sync.c (sync_start_init, sync_callbacks_load):
* exec/vsf_ykd.c (ykd_state_init):
* lcr/uis.c (cmd1):
* services/pload.c (send_message, token_callback, start_mcasting):
Here's another patch series to remove more warnings.
108 0001-add-void-parameter-list-to-avoid-isn-t-prototype.patch
56 0002-timer.c-connect-with-its-published-prototypes.patch
24 0003--exec-apidef.c-Include-apidef.h.patch
400 0004-declare-many-functions-static-also-avoids-missing.patch
From: Jim Meyering
* exec/timer.c: Include "timer.h".
* exec/timer.h (corosync_timer_init): Make return type in prototype
match the one in the function definition.
---
exec/timer.c |1 +
exec/timer.h |6 +++---
2 files changed, 4 insertions(+), 3 deletions(-)
diff --git a/exec/timer.c b
From: Jim Meyering
---
exec/apidef.c |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/exec/apidef.c b/exec/apidef.c
index 951a6a1..7148042 100644
--- a/exec/apidef.c
+++ b/exec/apidef.c
@@ -52,6 +52,7 @@
#include
#include "service.h"
#include
+#include "apidef.h"
L
From: Jim Meyering
* exec/objdb.c (objdb_wrlock, objdb_rdlock, objdb_rdunlock):
(objdb_wrunlock):
* services/cfg.c (send_shutdown, check_shutdown_status):
* services/votequorum.c (send_expectedvotes_notification):
* tools/corosync-cfgtool.c (shutdown_do):
---
exec/objdb.c |8
From: Jim Meyering
* tools/corosync-fplay.c (sync_printer_nada): Remove function.
* tools/corosync-objctl.c (get_child_name): Remove function.
* test/evsverify.c (msg): Remove unused variable.
---
test/evsverify.c|2 --
tools/corosync-fplay.c |4
tools/corosync-objctl.c |
From: Jim Meyering
* tools/corosync-fplay.c (records_printed): Remove.
---
tools/corosync-fplay.c |2 --
1 files changed, 0 insertions(+), 2 deletions(-)
diff --git a/tools/corosync-fplay.c b/tools/corosync-fplay.c
index f261c55..9eed9f7 100644
--- a/tools/corosync-fplay.c
+++ b/tools/coros
Three patches below:
Subject: [PATCH 1/3] cfg.h: add const
Subject: [PATCH 2/3] cfg.h: adjust parameter types: const+s/int/size_t/
Subject: [PATCH 3/3] cfg.h: mark "address_length" as an unused member
Regarding the unused "address_length" member, I suspected
that it should have type socklen
Jim Meyering wrote:
> Hello,
>
> [in corosync]
> Looking at a "const" cast-away warning related to an API issue,
> I wondered if cfg.c (corosync_cfg_state_track)'s notification_buffer
> parameter should *not* be "const", but do not see any use of that
> member, so can't be sure. No use in cluster
55 matches
Mail list logo
