- Description has changed:
Diff:
--- old
+++ new
@@ -8,6 +8,6 @@
May 13 02:31:53 sc2 osafimmnd[6042]: WA immnd_evt_proc_imm_init: PID 0
(7150) for 2020fb1d13cb8, MDS problem?
May 13 02:31:53 sc2 osafimmnd[6042]: WA immnd_evt_proc_imm_init: PID 0
(7150) for 2020fb1d13cb8, MDS
- Attachments has changed:
Diff:
--- old
+++ new
@@ -1 +0,0 @@
-sc2_mds.log (2.6 MB; text/x-log)
---
** [tickets:#1825] MDS: When saImmOmInitialize was invoked, MDS sent ‘pid = 0’
in MDS_CALLBACK_RECEIVE_INFO.**
**Status:** fixed
**Milestone:** 4.7.2
**Created:** Fri May 13,
- **status**: assigned --> review
---
** [tickets:#1825] MDS: When saImmOmInitialize was invoked, MDS sent ‘pid = 0’
in MDS_CALLBACK_RECEIVE_INFO.**
**Status:** review
**Milestone:** 4.7.2
**Created:** Fri May 13, 2016 12:34 PM UTC by Rafael
**Last Updated:** Thu May 26, 2016 04:46 AM UTC
Here's the result
X | PID exist | PID not exist
--- | --- | ---
OpenSAF as non-root | kill() returns -1errno is EPERM | kill() returns
-1errno is ESRCH
OpenSAF as root | kill() returns 0 | kill() returns -1errno is
ESRCH
---
** [tickets:#1825] MDS: When saImmOmInitialize was invoked,
Let me understand bit more about the return values in your environment before
publishing the final patch.
1) what dose`kill(info->pid, 0)` returning if PUD exist Opensaf as
NON-ROOT
2) what dose `kill(info->pid, 0)` returningif PUD NOT exist Opensaf as
NON-ROOT
3) what is the
Whic one works :
if ((info != NULL) && (kill(info->pid, 0) == -1) && (errno != EPERM))
or
if ((info != NULL) && (kill(info->pid, 0) == -1) && (errno == EPERM))
-AVM
---
** [tickets:#1825] MDS: When saImmOmInitialize was invoked, MDS sent ‘pid = 0’
in MDS_CALLBACK_RECEIVE_INFO.**
Then give a try with `if ((info != NULL) && ((kill(info->pid, 0) == -1) ||
errno == EPERM)`
in osaf/libs/core/mds/mds_dt_common.c and let me know the result
By the way with Linux flavor you are using ?
Are you running opensaf as NON root user ?
---
** [tickets:#1825] MDS: When
- **status**: unassigned --> assigned
- **assigned_to**: A V Mahesh (AVM)
- **Comment**:
> Hung Nguyen wrote:
>
> But sometimes we receive in order like this:
> Receive IMMND_EVT_A2ND_IMM_FINALIZE
> MDS_REGISTER_REQ messages
> MDS_UNREGISTER_REQ messages
> Receive
>>The only thing we can make sure that will come in order is NCSMDS_DOWN and
>>NCSMDS_UP.
>>So I still think that MDS should clear the down timer when receiving
>>NCSMDS_UP.
You mean NCSMDS_DOWN and NCSMDS_UP MDS events are kept on separate sockets
(Dsock and BSRsock) ?
or
Update:
I checked the log again and saw that the IMMA down event (NCSMDS_DOWN) is not
unexpected.
It just came late.
~~~
11:26:21.775374 [immsv_evt.c:5473] T8 Received: IMMND_EVT_A2ND_IMM_FINALIZE (5)
from 2010f
11:26:21.775529 [mds_main.c:0153] TR MDS: received 79 from 2010ffc93cf53, pid
Hi Mahesh,
Here's the immnd trace and mds log for this problem.
~~~
11:26:21.778367 osafimmnd[5238] NOTIFY |MDTM: svc down event for svc_id =
IMMA_OM(26), subscri. by svc_id = IMMND(25) pwe_id=1 Adest =
11:26:21.778474 osafimmnd[5238] NOTIFY |MDTM: svc up event
The test-case (simplified);
1. Start a virtual cluster
2. Installs the SW-under-test with a "campaign" with cluster reboot ~02:19, SAF
started ~02:20:09
3. Performs some tests (most likely not related)
4. Starts a "scale-out" by starting 2 new VM's in the cluster ~02:30
5. The new VM's
checked MDS logs :
IMMND PID [6042] started at time stamp of `May 13 2:20:11.902347`
May 13 2:20:11.902347 osafimmnd[6042] NOTIFY |BEGIN MDS LOGGING|
PID= | ARCHW=9|64bit=1
---
** [tickets:#1825] MDS: When saImmOmInitialize was invoked, MDS sent ‘pid = 0’
in MDS_CALLBACK_RECEIVE_INFO.**
**Status:** unassigned
**Milestone:** 4.7.2
**Created:** Fri May 13, 2016 12:34 PM UTC by Rafael
**Last Updated:** Fri May 13, 2016 12:34 PM UTC
**Owner:** nobody
14 matches
Mail list logo