- Description has changed:
Diff:
~~~~
--- old
+++ new
@@ -1,4 +1,4 @@
-Coredump happen when race condition between remove client thread and get
client thread in ntfd
+Coredump happens when try to access to a null pointer in ntfd.
~~~
BT ###
#0 0x0000558e5506a000 in NtfClient::getMdsDest() const ()
@@ -10,6 +10,7 @@
#6 0x0000558e55056637 in main ()
~~~
Step Reproduce:
-Step 1: send many notification from remote note to node active
-Step 2: toggling SIGSTOP/SIGCONT osaflogd process on node active, ntf's buffer
will be full.
-Step 3: reboot remote node.
+ 1. Simulate nfs hang to make the ntf buffer full.
+ 2. Client send many notifications to server.
+ 3. Reboot all client to simulate client down while sending ntf.
+ * Recommend: using a cluster with multiple node payloads to simulate
numerous clients sending notifications simultaneously. This will increase the
likelihood of overload, thereby raising the chances of encountering a core dump.
~~~~
---
**[tickets:#3359] ntf: fix ntfd coredump**
**Status:** assigned
**Milestone:** 5.24.09
**Created:** Thu Sep 05, 2024 07:25 AM UTC by Khuong Ba Le
**Last Updated:** Thu Sep 05, 2024 07:25 AM UTC
**Owner:** Khuong Ba Le
Coredump happens when try to access to a null pointer in ntfd.
~~~
BT ###
#0 0x0000558e5506a000 in NtfClient::getMdsDest() const ()
#1 0x0000558e55072213 in NtfAdmin::processNotification(unsigned int,
SaNtfNotificationTypeT, ntfsv_send_not_req*, mds_sync_snd_ctxt*, unsigned long
long) ()
#2 0x0000558e55072487 in NtfAdmin::notificationReceived(unsigned int,
SaNtfNotificationTypeT, ntfsv_send_not_req*, mds_sync_snd_ctxt*) ()
#3 0x0000558e5505805f in ?? ()
#4 0x0000558e55058844 in ?? ()
#5 0x0000558e55058c2b in ntfs_process_mbx ()
#6 0x0000558e55056637 in main ()
~~~
Step Reproduce:
1. Simulate nfs hang to make the ntf buffer full.
2. Client send many notifications to server.
3. Reboot all client to simulate client down while sending ntf.
* Recommend: using a cluster with multiple node payloads to simulate numerous
clients sending notifications simultaneously. This will increase the likelihood
of overload, thereby raising the chances of encountering a core dump.
---
Sent from sourceforge.net because [email protected] is
subscribed to https://sourceforge.net/p/opensaf/tickets/
To unsubscribe from further messages, a project admin can change settings at
https://sourceforge.net/p/opensaf/admin/tickets/options. Or, if this is a
mailing list, you can unsubscribe from the mailing list._______________________________________________
Opensaf-tickets mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensaf-tickets
