[IQUG] Failing in killing a given production IQ connection handle!

Ron Watkins rwatkins at dssolutions.com
Tue Jul 17 09:39:19 MST 2018


I have seen this behavior for many years. Basically, ive been told that the connection is executing a part of the code where an interrupt is either not accepted by the server or it’s not possible to interrupt the code at that location for some reason. In the past, IQ engineering has been able to tighten up some of these situations and get interrupts to work, but it does seem to be an on-going issue which (based on past history) will likely continue to occur going forward.

Note that if you do not collect the pstacks and other output Mark has suggested, they will not be able to close the gaps in interrupt handing, so it’s desirable going forward to continue to report as many of these situations as possible to gain ground on this problem. Note that if you have it hanging currently, there is a likelihood that it will re-occur if that same connection execution happens again. This is why reporting and fixing the issue will reduce the probability of it happening to you again.

Ron

 

From: iqug-bounces at iqug.org [mailto:iqug-bounces at iqug.org] On Behalf Of Steve Shen
Sent: Tuesday, July 17, 2018 9:32 AM
To: 'Mumy, Mark'; 'Mark Mumy'
Cc: iqug at iqug.org
Subject: Re: [IQUG] Failing in killing a given production IQ connection handle!

 

Mark,

 

I prefer to not put efforts in collecting those information because it’s IQ version 15.4. 

 

Multiple DBAs  ran into the same issues based on my Google results.  I did not believe that it was properly addressed and resolved by SAP/Sybase because the version was EOL.

 

I would rather focus my efforts more on my migration to version 16.x

 

Thank you.

 

Regards,

Steve 

From: Mumy, Mark [mailto:mark.mumy at sap.com] 
Sent: Tuesday, July 17, 2018 12:14 PM
To: Steve Shen; 'Mark Mumy'
Cc: iqug at iqug.org
Subject: Re: Failing in killing a given production IQ connection handle!

 

Yeah, just keep an eye out on your space usage and be ready to add more if needed.

 

If you can, grab the info that I put below.  The more data you can collect and get for support, the better off for us to figure out why it isn’t going away.

 

Mark

 

Mark Mumy

Strategic Technology Incubation Group

SAP Global Center of Excellence |  SAP

M +1 347-820-2136 | E  <mailto:mark.mumy at sap.com> mark.mumy at sap.com

My Blogs:  <https://blogs.sap.com/author/markmumy/> https://blogs.sap.com/author/markmumy/

                      

 <https://sap.na.pgiconnect.com/I825063> https://sap.na.pgiconnect.com/I825063

Conference tel: 18663127353,,8035340905#

 

From: Steve Shen <sshen at sscinc.com>
Date: Tuesday, July 17, 2018 at 10:45 AM
To: "markdmumy at gmail.com" <markdmumy at gmail.com>
Cc: "iqug at iqug.org" <iqug at iqug.org>, Mark Mumy <mark.mumy at sap.com>
Subject: RE: Failing in killing a given production IQ connection handle!

 

Mark,

 

I am  strongly hoping that we will be able to hold out the production applications until this Saturday.

 

Thank you.

 

Regards,

Steve Shen

From: Mark Mumy [mailto:markdmumy at gmail.com] 
Sent: Tuesday, July 17, 2018 11:28 AM
To: Steve Shen
Cc: iqug at iqug.org; Mark Mumy
Subject: Re: Failing in killing a given production IQ connection handle!

 

Steve,

 

Unfortunately, there’s not much you can do if the connection won’t drop.  I would recommend grabbing stack traces (pstack?) every 1-2 minutes for 3-4 iterations.  Then open a case and attach the pstacks.  Make sure to include things like sp_iqconnection, sp_iqversionuse, sp_iqcontext, etc.  This will help us figure out why it isn’t being killed.  More of a root cause analysis then helping the current situation.

 

The only way to remedy it if it won’t die is to reboot IQ.  But make sure to grab the above info so that we can figure out why and address it in the code.

 

Mark

 

On Jul 17, 2018, at 9:23 AM, Steve Shen <sshen at sscinc.com> wrote:

 

Today’s Topic:

 

1.       I failed in killing a connection handle from a critical IQ instance!

 

I killed a connection handle; but it was still there.

 

I killed the same connection handle multiple times, and it’s still there.  It has been for more than two hours.

 

I observed that the Other Version rocketing from 60 GB to more than 110 GB now. 

 

I failed in killing the connection handle for the first time ever!  The IQ performance is still good.

 

I am not allowed to shut down the production IQ.  

 

Is there any other  undocumented way to force killing this connection handle?

 

Please advise.  Thank you.

 

Regards,

Steve Shen



 

This email with all information contained herein or attached hereto may contain confidential and/or privileged information intended for the addressee(s) only. If you have received this email in error, please contact the sender and immediately delete this email in its entirety and any attachments thereto.

 

This email with all information contained herein or attached hereto may contain confidential and/or privileged information intended for the addressee(s) only. If you have received this email in error, please contact the sender and immediately delete this email in its entirety and any attachments thereto. 

This email with all information contained herein or attached hereto may contain confidential and/or privileged information intended for the addressee(s) only. If you have received this email in error, please contact the sender and immediately delete this email in its entirety and any attachments thereto. 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://iqug.org/pipermail/iqug/attachments/20180717/438a9326/attachment-0001.html>


More information about the IQUG mailing list