Restarting CBORD Interface Service Causes ‘error’ State

[Vince Angulo]

I didn’t see anything searching CBORD topics, so I’m looking to see if anyone’s seen this behavior before.

CBORD’s first response was to ask if this was a Cloverleaf issue, but I’ve never seen this happen with another interface.  

The one exception I can think of is if the thread is configured to go to a DNS address instead of an IP and there’s an interruption of any kind (and we don’t use DNS with our PROD connections).

If a third party connection goes offline, my experience is Cloverleaf goes to ‘opening’ and attempts to send until the other side resumes listening and we go back to ‘up’.  

I tend to think there might be something with how CBORD disconnects, but I’m not terribly savvy at what happens at the protocol level…

I’m letting our Operators know to stop and start the thread when this happens, but I can’t help but wonder if this should be happening at all.

Thanks in advance!

Topic

[Vince Angulo]

Might have found my own answer to this.  In our non-prods we leave Auto Connect off, and turn it on after boxing to prod.  I left it off.  Hopefully this corrects it.

[Jim Kosloskey]

What does the Process Log say?

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Vince Angulo]

I think this is the right section of the log.  I need to get together with the apps folks and do a focused test on the non-prod, cycling the interface service with auto-connect off then on.

[cmd :cmd :INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21] Receiving a command

[cmd :cmd :INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21] Received command: ‘. die’

[cmd :cmd :INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21] Doing ‘die’ command with args ”

[prod:prod:INFO/0:NSS_ORM_xlate:11/01/2018 15:01:21] Checking for leaked handles in the Xlate interpreter…

[tcl :out :INFO/0:NSS_ORM_xlate:11/01/2018 15:01:21]

[tcl :out :INFO/0:NSS_ORM_xlate:11/01/2018 15:01:21]

[prod:prod:INFO/0:NSS_ORM_xlate:11/01/2018 15:01:21] Checking for leaked handles in the General interpreter…

[tcl :out :INFO/0:NSS_ORM_xlate:11/01/2018 15:01:21]

[tcl :out :INFO/0:NSS_ORM_xlate:11/01/2018 15:01:21]

[prod:prod:INFO/0:  NSS_ORM_OUT:11/01/2018 15:01:21] Checking for leaked handles in the General interpreter…

[tcl :out :INFO/0:  NSS_ORM_OUT:11/01/2018 15:01:21]

[tcl :out :INFO/0:  NSS_ORM_OUT:11/01/2018 15:01:21]

[prod:prod:INFO/0:  NSS_ORM_OUT:11/01/2018 15:01:21] Checking for leaked handles in the TPS interpreter…

[tcl :out :INFO/0:  NSS_ORM_OUT:11/01/2018 15:01:21]

[tcl :out :INFO/0:  NSS_ORM_OUT:11/01/2018 15:01:21]

[sms :sms :ERR /0: NSS_ORM_PASS:11/01/2018 15:01:21] Tcl error:

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–] msgId = none

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–] proc = ‘overMsg’

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–] args = ”

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–] result = ”

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–] errorInfo: ‘

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–] Unknown mode ‘shutdown’ in overMsg

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–]     while executing

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–] “error “Unknown mode ‘$mode’ in overMsg””

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–]     (procedure “overMsg” line 25)

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–]     invoked from within

[sms :sms :ERR /0: NSS_ORM_PASS:–/–/—- –:–:–] “overMsg {CONTEXT sms_ob_data} {ARGS {}} {MODE shutdown} {VERSION 3.0}”‘

[prod:prod:INFO/0: NSS_ORM_PASS:11/01/2018 15:01:21] Checking for leaked handles in the General interpreter…

[tcl :out :INFO/0: NSS_ORM_PASS:11/01/2018 15:01:21]

[tcl :out :INFO/0: NSS_ORM_PASS:11/01/2018 15:01:21]

[prod:prod:INFO/0: NSS_ORM_PASS:11/01/2018 15:01:21] Checking for leaked handles in the TPS interpreter…

[tcl :out :INFO/0: NSS_ORM_PASS:11/01/2018 15:01:21]

[tcl :out :INFO/0: NSS_ORM_PASS:11/01/2018 15:01:21]

[cmd :cmd :INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21] Shutting down command thread NSS_ORM_cmd.

[prod:prod:INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21] Checking for leaked handles in the NCI interpreter…

[tcl :out :INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21]

[tcl :out :INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21]

[prod:prod:INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21] Checking for leaked handles in the General interpreter…

[tcl :out :INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21]

[tcl :out :INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21]

[cmd :cmd :INFO/0:  NSS_ORM_cmd:11/01/2018 15:01:21] Removing engine command port file.

[prod:prod:INFO/0:  STARTUP_TID:11/01/2018 15:01:21] Removing pid file.

[prod:prod:INFO/0:  STARTUP_TID:11/01/2018 15:01:21] Engine process 42532890 is terminating

[prod:prod:INFO/0:  STARTUP_TID:11/01/2018 15:01:21] Ended at Thu Nov  1 15:01:21 2018

[Jim Kosloskey]

Could it be the Tcl error is causing the issue? perhaps you have already researched that.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Vince Angulo]

I don’t think so.  We see that routinely, personally, I don’t think we have the overMsg proc properly implemented between our ‘pass’  and outbound threads — it fires every time the target ‘out’ thread shuts down.  Our level 2 programmer brought it back from a conference, and says that’s just the way it works.

Thanks for the suggestions though!

[Jim Kosloskey]

So I understand, the thread goes into an ‘Error’ state when you stop then start the thread?

Or is some other thing happening?

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Vince Angulo]

Okay, we did some regressive testing, checking one setting at a time, and found that the Auto Reconnect setting on the TCP/IP Configuration tab needs to be checked to allow Cloverleaf to remain in a listening state when the CBORD interface service bounced.

Thanks to our CBORD interface engineer, Jessica Smith, for checking with other Cloverleaf clients to see how they solved this issue.

[Jim Kosloskey]

Is that thread configured as a Client or Server on Cloverleaf?

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Vince Angulo]

Hi Jim,

It’s an outbound client thread.

[Author]

Replies