From: jjzhang <jjzhang.linux@gmail.com>
To: cluster-devel.redhat.com
Subject: [Cluster-devel] Re: [RFC PATCH] dlm: enhancing dlm_controld (pcmk) to be able to handle redundant rings
Date: Fri, 23 Oct 2009 18:38:25 +0800 [thread overview]
Message-ID: <20091023103645.GA6438@linux-jjzhang> (raw)
<andrew@beekhof.net>, Lars Marowsky-Bree <lmb@suse.de>
Cc: cluster-devel at redhat.com
Bcc:
Subject: Re: [Cluster-devel] Re: [RFC PATCH] dlm: enhancing dlm_controld
(pcmk) to be able to handle redundant rings
Reply-To:
In-Reply-To: <20091016101722.GZ15975@suse.de>
On Fri, Oct 16, 2009 at 12:17:22PM +0200, Lars Marowsky-Bree wrote:
> On 2009-10-15T09:51:08, David Teigland <teigland@redhat.com> wrote:
>
> > On Thu, Oct 15, 2009 at 09:43:56AM +0200, Andrew Beekhof wrote:
> > > Looks good to me :-)
> > > Although it might be nice to have the automated detection in place too.
> > >
> > > Less things for the admin to get wrong. What do you think Dave?
> >
> > I don't mind, now that I've taken a moment to think how it should work...
> > There should be three protocol options, "tcp", "sctp", "detect". Detect
> > should work by reading the totem/rrp_mode value in setup_ccs().
>
> Can we then also agree on "detect" being the default?
>
> (Experience shows that admins will misconfigure just about anything if
> they have to deviate from the defaults.)
Hello,
Below is the improved patch. I have set the "default" as the same logic as
"detect", but I'm not sure if this is OK.
Thanks a lot for your kindly review and comments :)
Signed-off-by: Jiaju Zhang <jjzhang.linux@gmail.com>
---
configure.ac | 1 +
group/dlm_controld/Makefile.am | 4 ++-
group/dlm_controld/main.c | 9 ++++++-
group/dlm_controld/pacemaker.c | 48 ++++++++++++++++++++++++++++++++++++++++
4 files changed, 60 insertions(+), 2 deletions(-)
diff --git a/configure.ac b/configure.ac
index d50de87..e0a1a0d 100644
--- a/configure.ac
+++ b/configure.ac
@@ -131,6 +131,7 @@ if test "x${enable_pacemaker}" = xyes; then
PKG_CHECK_MODULES([totempg],[libtotem_pg])
PKG_CHECK_MODULES([xml],[libxml-2.0])
PKG_CHECK_MODULES([glib],[glib-2.0])
+ PKG_CHECK_MODULES([cfg],[libcfg])
check_lib_no_libs cib cib_new
check_lib_no_libs crmcluster crm_set_status_callback
check_lib_no_libs crmcommon init_server_ipc_comms
diff --git a/group/dlm_controld/Makefile.am b/group/dlm_controld/Makefile.am
index c14ab89..ce6e587 100644
--- a/group/dlm_controld/Makefile.am
+++ b/group/dlm_controld/Makefile.am
@@ -62,12 +62,14 @@ dlm_controld_pcmk_CPPFLAGS= $(shared_CPPFLAGS) \
dlm_controld_pcmk_CFLAGS = $(shared_CFLAGS) \
$(glib_CFLAGS) \
$(xml_CFLAGS) \
- $(totempg_CFLAGS)
+ $(totempg_CFLAGS) \
+ $(cfg_CFLAGS)
dlm_controld_pcmk_LDFLAGS = $(shared_LIBS) \
$(glib_LIBS) \
$(xml_LIBS) \
$(totempg_LIBS) \
+ $(cfg_LIBS) \
-lcib -lcrmcommon -lcrmcluster
dlm_controld_pcmk_LDADD = $(shared_LDADD)
diff --git a/group/dlm_controld/main.c b/group/dlm_controld/main.c
index f90cd21..8949d11 100644
--- a/group/dlm_controld/main.c
+++ b/group/dlm_controld/main.c
@@ -1058,11 +1058,13 @@ static void print_usage(void)
printf(" Default is %u\n", DEFAULT_DROP_RESOURCES_COUNT);
printf(" -a <ms> plock ownership drop resources age (milliseconds)\n");
printf(" Default is %u\n", DEFAULT_DROP_RESOURCES_AGE);
+ printf(" -r <num> DLM in-kernel communication protocols: TCP(0), SCTP(1), DETECT(2)\n");
+ printf(" Default is auto-detect (2)");
printf(" -h Print this help, then exit\n");
printf(" -V Print program version information, then exit\n");
}
-#define OPTION_STRING "LDKf:q:d:p:Pl:o:t:c:a:hV"
+#define OPTION_STRING "LDKf:q:d:p:Pl:o:t:c:a:r:hV"
static void read_arguments(int argc, char **argv)
{
@@ -1142,6 +1144,11 @@ static void read_arguments(int argc, char **argv)
cfgd_drop_resources_age = atoi(optarg);
break;
+ case 'r':
+ optk_protocol = 1;
+ cfgk_protocol = atoi(optarg);
+ break;
+
case 'h':
print_usage();
exit(EXIT_SUCCESS);
diff --git a/group/dlm_controld/pacemaker.c b/group/dlm_controld/pacemaker.c
index 810c644..b3b147a 100644
--- a/group/dlm_controld/pacemaker.c
+++ b/group/dlm_controld/pacemaker.c
@@ -22,7 +22,42 @@
#include <pacemaker/crm/msg_xml.h>
#include <pacemaker/crm/cib.h>
+#include <corosync/corotypes.h>
+#include <corosync/cfg.h>
+
#define COMMS_DIR "/sys/kernel/config/dlm/cluster/comms"
+#define PROTO_TCP 0
+#define PROTO_SCTP 1
+#define PROTO_DETECT 2
+
+static int get_ring_count(void)
+{
+ cs_error_t result;
+ corosync_cfg_handle_t handle;
+ unsigned int interface_count;
+ char **interface_names;
+ char **interface_status;
+
+ result = corosync_cfg_initialize(&handle, NULL);
+ if (result != CS_OK) {
+ log_error("Failed to initialize corosync configuration API (error=%d)", result);
+ (void)corosync_cfg_finalize(handle);
+ return -1;
+ }
+
+ result = corosync_cfg_ring_status_get(handle,
+ &interface_names,
+ &interface_status,
+ &interface_count);
+ if (result != CS_OK) {
+ log_error("Failed to get the ring status (error=%d)", result);
+ (void)corosync_cfg_finalize(handle);
+ return -1;
+ }
+
+ (void)corosync_cfg_finalize(handle);
+ return interface_count;
+}
int setup_ccs(void)
{
@@ -30,6 +65,19 @@ int setup_ccs(void)
* only allow configuration from the command-line until CoroSync is stable
* enough to be used with Pacemaker
*/
+ if (cfgk_protocol != -1 && cfgk_protocol != PROTO_TCP &&
+ cfgk_protocol != PROTO_SCTP && cfgk_protocol != PROTO_DETECT) {
+ log_error("ignore invalid value %d for dlm protocol", cfgk_protocol);
+ cfgk_protocol = -1;
+ }
+
+ if (cfgk_protocol == PROTO_DETECT || cfgk_protocol == -1) {
+ if (get_ring_count() > 1)
+ cfgk_protocol = PROTO_SCTP;
+ else
+ cfgk_protocol = PROTO_TCP;
+ }
+
return 0;
}
next reply other threads:[~2009-10-23 10:38 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-23 10:38 jjzhang [this message]
2009-10-23 13:05 ` [Cluster-devel] Re: [RFC PATCH] dlm: enhancing dlm_controld (pcmk) to be able to handle redundant rings Andrew Beekhof
2009-10-23 13:23 ` Jiaju Zhang
2009-10-23 13:34 ` Lars Marowsky-Bree
2009-10-23 17:55 ` David Teigland
2009-10-23 19:18 ` Steven Dake
2009-10-23 20:51 ` David Teigland
2009-10-26 10:44 ` Jiaju Zhang
-- strict thread matches above, loose matches on Subject: below --
2009-10-14 10:50 [Cluster-devel] " Jiaju Zhang
2009-10-14 13:28 ` [Cluster-devel] " Jiaju Zhang
2009-10-14 18:17 ` David Teigland
2009-10-15 5:34 ` Jiaju Zhang
2009-10-15 7:43 ` Andrew Beekhof
2009-10-15 14:51 ` David Teigland
2009-10-15 15:30 ` Andrew Beekhof
2009-10-15 16:07 ` Jiaju Zhang
2009-10-16 10:17 ` Lars Marowsky-Bree
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20091023103645.GA6438@linux-jjzhang \
--to=jjzhang.linux@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).