diff mbox series

main/process: Add extra sockname debugging

Message ID 20221213232502.114660-1-richard.purdie@linuxfoundation.org
State Accepted, archived
Commit 376a516dc8c96727fd042ada65f803013601ee2d
Headers show
Series main/process: Add extra sockname debugging | expand

Commit Message

Richard Purdie Dec. 13, 2022, 11:25 p.m. UTC 
We're struggling to understand how bitbake.sock can sometimes disappear
in live builds when we can't see where it could have been deleted.
This causes connection failures to the server and failed builds.

Add some extra debugging around the server log and client retry
log messages to give more information for the next time this issue
occurs.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
---
 lib/bb/main.py           | 3 ++-
 lib/bb/server/process.py | 7 ++++---
 2 files changed, 6 insertions(+), 4 deletions(-)
diff mbox series

Patch

diff --git a/lib/bb/main.py b/lib/bb/main.py
index ee12256bc8..ed3e1ede61 100755
--- a/lib/bb/main.py
+++ b/lib/bb/main.py
@@ -446,7 +446,8 @@  def setup_bitbake(configParams, extrafeatures=None):
                         logger.info("Previous bitbake instance shutting down?, waiting to retry... (%s)" % timestamp())
                         procs = bb.server.process.get_lockfile_process_msg(lockfile)
                         if procs:
-                            logger.info("Processes holding bitbake.lock:\n%s" % procs)
+                            logger.info("Processes holding bitbake.lock (missing socket %s):\n%s" % (sockname, procs))
+                        logger.info("Directory listing: %s" % (str(os.listdir(topdir))))
                         i = 0
                         lock = None
                         # Wait for 5s or until we can get the lock
diff --git a/lib/bb/server/process.py b/lib/bb/server/process.py
index f4ab80ba67..44c65451fc 100644
--- a/lib/bb/server/process.py
+++ b/lib/bb/server/process.py
@@ -154,9 +154,10 @@  class ProcessServer():
             fds.append(self.xmlrpc)
         seendata = False
         serverlog("Entering server connection loop")
+        serverlog("Lockfile is: %s\nSocket is %s (%s)" % (self.bitbake_lock_name, self.sockname, os.path.exists(self.sockname)))
 
         def disconnect_client(self, fds):
-            serverlog("Disconnecting Client")
+            serverlog("Disconnecting Client (socket: %s)" % os.path.exists(self.sockname))
             if self.controllersock:
                 fds.remove(self.controllersock)
                 self.controllersock.close()
@@ -246,7 +247,7 @@  class ProcessServer():
                 try:
                     serverlog("Running command %s" % command)
                     self.command_channel_reply.send(self.cooker.command.runCommand(command))
-                    serverlog("Command Completed")
+                    serverlog("Command Completed (socket: %s)" % os.path.exists(self.sockname))
                 except Exception as e:
                    stack = traceback.format_exc()
                    serverlog('Exception in server main event loop running command %s (%s)' % (command, stack))
@@ -273,7 +274,7 @@  class ProcessServer():
 
             ready = self.idle_commands(.1, fds)
 
-        serverlog("Exiting")
+        serverlog("Exiting (socket: %s)" % os.path.exists(self.sockname))
         # Remove the socket file so we don't get any more connections to avoid races
         try:
             os.unlink(self.sockname)