Can’t close a Process

This topic has 18 replies, 4 voices, and was last updated 19 years, 2 months ago by Anonymous.

Creator

Topic
October 3, 2005 at 5:36 pm #48065
Doug Stelley
Participant
No matter what I’ve tried I can’t seem to be able to close one of my Processes. within it are 3 inbound threads (all stuck on “Initializing” and an Outgoing(stuck on “Initializing”). These are HL7 TCP/IP. When I try to execute an”hcicmd -p meta_1 -c “InMetaRde pstop” I get the error:

“Unable to contach process “meta_1” on port 4175

No Connection could be made because the target machine actively refused it”.

“hcicmd failed”

“Now trying SIGINT”

This isn’t using port 4175 (or even close)

I just want to force it closed, I’ve even bounced the service, no good.

HELP!!!

Win2Ksp4, Qdx 5.3 rev3

Doug
Creator

Topic

Viewing 17 reply threads

Author

Replies
- October 3, 2005 at 5:52 pm #57513
  Doug Stelley
  Participant
  using HCITCPTEST to a one of the local ports gets this, How can I be refusing it?
  
  Doug
  
  ******************************
  
  Command Issued: hcitcptest -t 4 -h localhost -p 8630
  
  socket error: No connection could be made because the target machine actively refused it.
  
  connecting…
- October 3, 2005 at 7:51 pm #57514
  Charlie Bursell
  Participant
  This is a WINSOCK error : WSAEHOSTDOWN
  
  It simply means you are tring to connect to a server on a socket but that server is not listening on that port.
  
  You can verify this by running netstat -an and looking for that port to see if there is a listen on it.
  
  For the engine problem, it looks like the TCP/IP connection was lost. Stop the engine, not the thread: hcienginestop -p
  It may take a while as it may have to kick it to death.
- October 4, 2005 at 11:10 am #57515
  Doug Stelley
  Participant
  No Luck,
  
  Running netstat -an showed me not listening on that port.
  
  this is the response from running hcienginestop:
  
  C:quovadxqdx5.3integratorsupport>hcienginestop -p meta_1
  
  Trying hcicmd…
  
  Can’t contact process meta_1 on port 4175: No connection could be made because the target machine actively refused it.
  
  hcicmd failed!
  
  Now trying SIGINT…
  
  Unable to send SIGINT: No such file or directory
  
  HELP!!
- October 4, 2005 at 12:47 pm #57516
  Anonymous
  Participant
  Doug,
  
  I’m not familiar with Cloverleaf on Windows, but do you have a firewall or something similar running on the Windows box? If you do, try creating some rules to allow Cloverleaf to use any port.
  
  My guess for the first error is that the server is not up since is not connecting, but the second error (cannot bring down the process) is an indication of something stopping the socket communication (the command to stop the process is sent using a port) …that is why I think that there is a firewall or something stopping that type of communications.
  
  Some versions of Windows have a built in firewall, check if that is your problem.
- October 4, 2005 at 1:43 pm #57517
  Doug Stelley
  Participant
  There must be some file or something I can modify to force close a connection. I have to get this closed.
  
  I can recreate the threads and process if its deleted, no problem. But the ports are in use I need to listen on.
  
  Give me a command that will absolutely kill a process. I want to delete from existance. rebuild it from scratch and launch it, but i need it closed. Work has stopped in that department until i get something going HHEELLPP!!!
- October 4, 2005 at 1:50 pm #57518
  Anonymous
  Participant
  Doug,
  
  In UNIX we use the “kill” command. I’m not sure if there is something similar in Windows. You may have to go to the Task Manager and kill the process from there… or bounce the Windows box (probably faster)
  
  Regards,
  
  Carlos
- October 4, 2005 at 1:55 pm #57519
  Doug Stelley
  Participant
  I’ve Bounced the service in Windows, I’ve rebooted the server. I’ve run “hcienginestop” several times, but it always stays “up”
  
  the threads stay “initializing”.
  
  No Joy
- October 4, 2005 at 2:04 pm #57520
  Doug Stelley
  Participant
  What file is the “Command and Engine Output” copied to? I’d paste the actual responses if I can find it…
- October 4, 2005 at 2:05 pm #57521
  Anonymous
  Participant
  After rebooting the server the process is up? That is strange… Is Cloverleaf starting the process when the server comes up?
  
  Did you check for a firewall?
- October 4, 2005 at 2:08 pm #57522
  Anonymous
  Participant
  The engine output for the process goes to the process log in the process directory.
  
  site/exec/processes/processname/processname.log
- October 4, 2005 at 2:09 pm #57523
  Doug Stelley
  Participant
  The quovadx service is automatic to start. The process however is not. I have no idea why it shows as running. I can’t save it though using Network configurater as I get an error that the process needs to stop before saving…..
- October 4, 2005 at 2:11 pm #57524
  Doug Stelley
  Participant
  The quovadx service is automatic to start. The process however is not. I have no idea why it shows as running. I can’t save it though using Network configurater as I get an error that the process needs to stop before saving…..
  
  🙁 🙁 🙁
- October 4, 2005 at 2:15 pm #57525
  Anonymous
  Participant
  In Unix we have a file called “pid” under the process directory. If for some reason the process dies and the file is still there, Cloverleaf may get confused and think that the process is still up. Deleting that file may solve the problem
  
  ./site/exec/processes/proces_name/pid
  
  You may want to call Quovadx support if this is a production issue. They should be able to help you faster over the phone.
- October 4, 2005 at 2:24 pm #57526
  Doug Stelley
  Participant
  retyped from the command and engine output “command output”
  
  I did find the “4175” in the “cmd_port” within the meta_1 directory
  
  hcicmd -p meta_1 -c “META_IN,InMetaRde,META_OUT,InMetaMfn purgex”
  
  error was: No such file or directory
  
  Try to connect with OLEAN_CLOVERLEA
  
  Unable to contact process ‘meta_1’on port 4175
  
  Unable to contact process ‘meta_1’on port 4175
  
  Error was: No connection could be made because the target machine actively refused it.
  
  hcicmd -p meta_1 -c”META_IN,InMetaRde,META_OUT,InMetaMfn eo_alias enable_all”
  
  error was: No such file or directory
  
  Try to connect with OLEAN_CLOVERLEA
  
  Unable to contact process ‘meta_1’ on port 4175
  
  Unable to contact process ‘meta_1’ on port 4175
  
  Error was: No connection could be made because the target machine actively refused it.
  
  hcienginestop -p meta_1
  
  Trying hcicmd…
  
  Now trying SIGINT…
  
  Now trying SIGKILL…
  
  Unable to stop process ‘meta_1’!
  
  hcienginerun -p meta_1
  
  Engine process ‘meta_1’ is already running.
- October 4, 2005 at 2:30 pm #57527
  Doug Stelley
  Participant
  “pid” huh?
  
  OK, I deleted this file and finally it shows as “dead”
  
  sheesh.
  
  now to see if I can save anything. Thanks Carlos
- October 4, 2005 at 2:43 pm #57528
  Doug Stelley
  Participant
  I made a needed change and it was successful, all is “up” and in the green. Again Thanks 😀
- October 4, 2005 at 2:54 pm #57529
  garry r fisher
  Participant
  Hi Doug,
  
  I’ve seen this problem a number of times and appears to be due to the process not closing down properly – possibly a Windows thing. This leaves the pid file in the processes directory. I think all the GUI does is check for the presence of this file and if it finds it shows the process as ‘up’ even though it is definitely dead.
  
  I usually check for this file and the cmd_port (in the same directory) if I have this type of problem and delete them once I’ve confirmed using task manager that no hciengine is running for the process.
  
  Regards
  
  Garry
- October 10, 2005 at 2:05 pm #57530
  Anonymous
  Participant
  hi Doug,
  
  engine will not allow you to let connect to a port that is being referenced by a thread that went down abnormally.
  
  1. You ought to be very cautious on this. I am not familiar on QDX running Win2k.
  
  You need to clean up the engine.
  
  If the issue lies within only one process, you can do the following:
  
  hcilmclear -p
  This should cleanup the engine at the process level.
  
  2. In the netconfig, go to the thread you were having problems, disable the auto start option and save it.
  
  After the cleanup done in step 1, you should be able to start the threads manually.
  
  -Reggie-
Author

Replies

Viewing 17 reply threads

The forum ‘Cloverleaf’ is closed to new topics and replies.