Script 'mail_helper' called by obssrc
Hello community,

here is the log from the commit of package yast2-iscsi-client for 
openSUSE:Factory checked in at 2025-11-01 23:34:22
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Comparing /work/SRC/openSUSE:Factory/yast2-iscsi-client (Old)
 and      /work/SRC/openSUSE:Factory/.yast2-iscsi-client.new.1980 (New)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Package is "yast2-iscsi-client"

Sat Nov  1 23:34:22 2025 rev:153 rq:1314909 version:5.0.11

Changes:
--------
--- /work/SRC/openSUSE:Factory/yast2-iscsi-client/yast2-iscsi-client.changes    
2025-08-21 16:54:35.302081695 +0200
+++ 
/work/SRC/openSUSE:Factory/.yast2-iscsi-client.new.1980/yast2-iscsi-client.changes
  2025-11-01 23:34:24.565953199 +0100
@@ -1,0 +2,10 @@
+Mon Oct 20 10:16:23 UTC 2025 - Ancor Gonzalez Sosa <[email protected]>
+
+- Do not perform an extra discovery during iBFT login, to prevent
+  the creation of problematic records at the node, interface and fw
+  databases (bsc#1247711).
+- Introduce a delay after iBFT login, to increase the chances of
+  libstorage-ng to properly detect the new devices (bsc#1247711).
+- 5.0.11
+
+-------------------------------------------------------------------

Old:
----
  yast2-iscsi-client-5.0.10.tar.bz2

New:
----
  yast2-iscsi-client-5.0.11.tar.bz2

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Other differences:
------------------
++++++ yast2-iscsi-client.spec ++++++
--- /var/tmp/diff_new_pack.LI7yie/_old  2025-11-01 23:34:25.133976949 +0100
+++ /var/tmp/diff_new_pack.LI7yie/_new  2025-11-01 23:34:25.137977116 +0100
@@ -17,7 +17,7 @@
 
 
 Name:           yast2-iscsi-client
-Version:        5.0.10
+Version:        5.0.11
 Release:        0
 Summary:        YaST2 - iSCSI Client Configuration
 License:        GPL-2.0-only

++++++ yast2-iscsi-client-5.0.10.tar.bz2 -> yast2-iscsi-client-5.0.11.tar.bz2 
++++++
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn' 
'--exclude=.svnignore' 
old/yast2-iscsi-client-5.0.10/package/yast2-iscsi-client.changes 
new/yast2-iscsi-client-5.0.11/package/yast2-iscsi-client.changes
--- old/yast2-iscsi-client-5.0.10/package/yast2-iscsi-client.changes    
2025-08-20 17:26:00.000000000 +0200
+++ new/yast2-iscsi-client-5.0.11/package/yast2-iscsi-client.changes    
2025-10-31 15:22:36.000000000 +0100
@@ -1,4 +1,14 @@
 -------------------------------------------------------------------
+Mon Oct 20 10:16:23 UTC 2025 - Ancor Gonzalez Sosa <[email protected]>
+
+- Do not perform an extra discovery during iBFT login, to prevent
+  the creation of problematic records at the node, interface and fw
+  databases (bsc#1247711).
+- Introduce a delay after iBFT login, to increase the chances of
+  libstorage-ng to properly detect the new devices (bsc#1247711).
+- 5.0.11
+
+-------------------------------------------------------------------
 Tue Aug 19 10:52:39 UTC 2025 - Knut Anderssen <[email protected]>
 
 - Fix the initialization of the valid iscsi offload cards not 
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn' 
'--exclude=.svnignore' 
old/yast2-iscsi-client-5.0.10/package/yast2-iscsi-client.spec 
new/yast2-iscsi-client-5.0.11/package/yast2-iscsi-client.spec
--- old/yast2-iscsi-client-5.0.10/package/yast2-iscsi-client.spec       
2025-08-20 17:26:00.000000000 +0200
+++ new/yast2-iscsi-client-5.0.11/package/yast2-iscsi-client.spec       
2025-10-31 15:22:36.000000000 +0100
@@ -17,7 +17,7 @@
 
 
 Name:           yast2-iscsi-client
-Version:        5.0.10
+Version:        5.0.11
 Release:        0
 Summary:        YaST2 - iSCSI Client Configuration
 License:        GPL-2.0-only
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn' 
'--exclude=.svnignore' 
old/yast2-iscsi-client-5.0.10/src/modules/IscsiClientLib.rb 
new/yast2-iscsi-client-5.0.11/src/modules/IscsiClientLib.rb
--- old/yast2-iscsi-client-5.0.10/src/modules/IscsiClientLib.rb 2025-08-20 
17:26:00.000000000 +0200
+++ new/yast2-iscsi-client-5.0.11/src/modules/IscsiClientLib.rb 2025-10-31 
15:22:36.000000000 +0100
@@ -1046,13 +1046,11 @@
       if !getiBFT.empty?
         result = SCR.Execute(path(".target.bash_output"), GetAdmCmd("-m fw 
-l"))
         ret = false if result["exit"] != 0
-
-        # Note that fw discovery does not store persistent records in the node 
or discovery DB,
-        # so this is likely only done for reporting purposes (writing the info 
to the YaST logs)
         log.info "Autologin into iBFT : #{result}"
-        result = SCR.Execute(path(".target.bash_output"), GetAdmCmd("-m 
discovery -t fw"))
-        log.info "iBFT discovery: #{result}"
+        # Give iSCSI some time, even if logon was not 100% successful
+        autologon_delay
       end
+      log.info "end of autoLogOn function"
       ret
     end
 
@@ -1622,6 +1620,32 @@
 
       [dev_name, hwaddr, type_label].compact.join(" - ")
     end
+
+    # Method to be called after an iBFT login, to mitigate some known timing 
issues.
+    #
+    # See bsc#1247711 for some background.
+    #
+    # When this method is called, we are already logged into the iSCSI 
targets. But device detection
+    # in the kernel is asynchronous and the corresponding iSCSI devices may 
take some time to appear
+    # in the system. Moreover, there is no guarantee that the devices will all 
show up. According to
+    # a comment at bsc#1247711: "One or more paths being down is an acceptable 
situation. They
+    # may/should come up later, but there is no requirement that they be there 
at any given time".
+    #
+    # That could be a big problem for yast-storage-ng, which expects a 
complete and stable set of
+    # devices during its probing phase (eg. to properly detect multipath).
+    def autologon_delay
+      log.info "Giving some time to iSCSI"
+      sleep(10)
+      # Execute the following command just for logging purposes, it shows the 
relationship between
+      # existing sessions and kernel device names. Theoretically it could be 
used to detect whether
+      # all devices are available before the hard-coded delay of the previous 
line, but we decided
+      # to keep the code simple.
+      res = SCR.Execute(path(".target.bash_output"), GetAdmCmd("-m session 
-P3"))
+      log.info "iscsiadm result: #{res}"
+      # Invoking udevadm just in case some of the devices showed up but it is 
not ready yet.
+      # This probably makes no difference, but it will not hurt either.
+      Yast::Execute.locally("/usr/bin/udevadm", "settle", "--timeout=15")
+    end
   end
 
   IscsiClientLib = IscsiClientLibClass.new

Reply via email to