Re: VTAM session information - rewording of an earlier question.
Richard, I'll try to provide more details. We have historically been a RUMBA shop, using RUMBA for normal telnet traffic going to our z9. We also have an application that uses RUMBA HLLAPI interfaces for screen scraping into other software packages. Recently management decided they wanted to switch our emulator from RUMBA to HOD. (political decision that I can't do anything about) Anyway, once the developers started messing with HOD, we have occasionally had the situation where a block of LUs will become unavailable to RUMBA. When this happens, the victim will get a black screen in RUMBA. The LUname will show on the RUMBA screen, but our TPX session manager main screen will fail to appear. The user will disconnect/reconnect, getting the next LUname and probably another black screen. Eventually they will get a session and TPX will come up. If, in the middle of this, the person tries to connect with HOD, they will get right in. I think it is a timing issue of some sort but haven't been able to track it down. Hence, I was hoping to get some kind of historical VTAM LU session start/end records and compare them (brute force) to the SMF 119 records out of TCPIP. Unfortunately I know nothing about HOD and the setup they used on it. Rex -Original Message- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Richard Peurifoy Sent: Tuesday, February 13, 2007 5:16 PM To: IBM-MAIN@BAMA.UA.EDU Subject: Re: VTAM session information - rewording of an earlier question. Pommier, Rex R. wrote: Unfortunately Netview isn't part of my software mix. At one time years ago, we had netspy from CA and I would assume it gave me that capability as well. I was hoping that VTAM would cut SMF records out-of-the-box, but it appears as though it doesn't. I'm sure mgmt would never allow me the time to build an exit to gather the information either. Thanks for the info. Can you provide more detail on your problem, maybe there is another way to look at it. -- Richard -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: VTAM session information - rewording of an earlier question.
Pommier, Rex R. wrote: Richard, I'll try to provide more details. We have historically been a RUMBA shop, using RUMBA for normal telnet traffic going to our z9. We also have an application that uses RUMBA HLLAPI interfaces for screen scraping into other software packages. Recently management decided they wanted to switch our emulator from RUMBA to HOD. (political decision that I can't do anything about) Anyway, once the developers started messing with HOD, we have occasionally had the situation where a block of LUs will become unavailable to RUMBA. When this happens, the victim will get a black screen in RUMBA. The LUname will show on the RUMBA screen, but our TPX session manager main screen will fail to appear. The user will disconnect/reconnect, getting the next LUname and probably another black screen. Eventually they will get a session and TPX will come up. If, in the middle of this, the person tries to connect with HOD, they will get right in. I think it is a timing issue of some sort but haven't been able to track it down. Hence, I was hoping to get some kind of historical VTAM LU session start/end records and compare them (brute force) to the SMF 119 records out of TCPIP. Unfortunately I know nothing about HOD and the setup they used on it. It sounds like the TCP LU is occasionally being left connected to some application and the RUMBA connection is landing in the middle of it. Do you have LOSTERM=IMMED coded on the VTAM appl definitions being used by TCP? Have you displayed the LU in VTAM when RUMBA gets a dead connection (D NET,ID=luname,E)? When the user disconnect does it still look the same? What does TCP show (D TCPIP,tcpjobname,T,CONN,LUN=luname) The TPX log also shows LOGON, SIGNON, SIGNOFF, and LOGOFF as well as TPX virtual terminal connect/disconnect. TPX can also be configured to log this to SMF. Is TPX your default appl in TCP, or are you using the TCP USS screens (or some other solicitor)? -- Richard -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
VTAM session information - rewording of an earlier question.
Is there anything in SMF or elsewhere that shows historical session start/end information from a VTAM-centric view? Using the TELNET server gives me the 118/119 SMF output from a TCP/IP view. I have the situation where it appears TCP/IP and VTAM aren't keeping their LU's in sync and it appears TCP/IP is releasing the LU and VTAM isn't. I can't find anything in SMF or VTAM documentation that shows VTAM cutting SMF or log records showing this kind of info. Thanks. Rex -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: VTAM session information - rewording of an earlier question.
Pommier, Rex R. wrote: Is there anything in SMF or elsewhere that shows historical session start/end information from a VTAM-centric view? Using the TELNET server gives me the 118/119 SMF output from a TCP/IP view. I have the situation where it appears TCP/IP and VTAM aren't keeping their LU's in sync and it appears TCP/IP is releasing the LU and VTAM isn't. I can't find anything in SMF or VTAM documentation that shows VTAM cutting SMF or log records showing this kind of info. Thanks. Rex If you have NETVIEW you can configure the NLDM component to write SMF records for session start and end (record 39 IIRC). Other products my do this as well. As a last resort you might be able to do this yourself in a VTAM session management exit. -- Richard -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: VTAM session information - rewording of an earlier question.
Unfortunately Netview isn't part of my software mix. At one time years ago, we had netspy from CA and I would assume it gave me that capability as well. I was hoping that VTAM would cut SMF records out-of-the-box, but it appears as though it doesn't. I'm sure mgmt would never allow me the time to build an exit to gather the information either. Thanks for the info. Rex -Original Message- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Richard Peurifoy Sent: Tuesday, February 13, 2007 4:05 PM To: IBM-MAIN@BAMA.UA.EDU Subject: Re: VTAM session information - rewording of an earlier question. Pommier, Rex R. wrote: Is there anything in SMF or elsewhere that shows historical session start/end information from a VTAM-centric view? Using the TELNET server gives me the 118/119 SMF output from a TCP/IP view. I have the situation where it appears TCP/IP and VTAM aren't keeping their LU's in sync and it appears TCP/IP is releasing the LU and VTAM isn't. I can't find anything in SMF or VTAM documentation that shows VTAM cutting SMF or log records showing this kind of info. Thanks. Rex If you have NETVIEW you can configure the NLDM component to write SMF records for session start and end (record 39 IIRC). Other products my do this as well. As a last resort you might be able to do this yourself in a VTAM session management exit. -- Richard -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: VTAM session information - rewording of an earlier question.
Pommier, Rex R. wrote: Unfortunately Netview isn't part of my software mix. At one time years ago, we had netspy from CA and I would assume it gave me that capability as well. I was hoping that VTAM would cut SMF records out-of-the-box, but it appears as though it doesn't. I'm sure mgmt would never allow me the time to build an exit to gather the information either. Thanks for the info. Can you provide more detail on your problem, maybe there is another way to look at it. -- Richard -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
