We are running RV on z/OS and saw the IEC223I message in the JOB LOG, what seems to be the problem?
Article ID: KB0090658
Updated On:
| Products | Versions |
|---|---|
| TIBCO Rendezvous | - |
| Not Applicable | - |
Description
Resolution:
The following IBM PTFs need to be applied that involves the IEC223I message.
UK02664
UK04264 (PRE UK01284)
UA18639
In one reported situation, simply applying PTF UK02664 eliminated the IEC223I message.
Important Note: this is not a TIBCO Rendezvous related problem; if you receive the IEC223I message, you should apply the above IBM PTFs on the mainframe machine.
The problem is a 'C' and LE run-time issue involving the function call fseek().
Additional information:
For the I/O code fixed by PQ96742 (PTF UK02664 for R709 = z/OS V1.6; and PTF UK02662 for R707 = z/OS V1.4), the module updated was EDCBP030 – this includes the sp30 code seen in the failing Traceback flow that contains:
“TIBCO” -> fseek -> sp30 -> osseek -> CEL4I031 -> osopenru -> OSBCLOSE
sp30 has code that maintains the amount of data in a data set, and it determines when a data set has been extended. In effect, it can determine when a multivol data set moved to another volume. This is what has happened here - the data set filled the first volume and moved to another volume. The code in sp30 could not handle this. sp30, upon determining that a data set had moved onto another volume invalidly *forced* a re-open of the data set (done by osopenru), so any new data that might have been written on the second volume was accounted for. This caused a close (OSBCLOSE) to be done as part of the re-open. The test to determine if a re-open is necessary, however, was bad. PQ96742 also fixed this test in sp30, so an invalid re-open is “not” done – and, the data on the new volume is simply part of the fseek (no re-open / OSBCLOSE necessary).
Therefore:
1) PQ96742 can lead to MSGIEC2231 and RC20 in certain environments
2) PQ96742 relates to data spanning volumes for a multi-volume data set. EDCBP030 was also fixed, so a re-open of a data set is not performed when volumes are crossed. This re-open was leading to an invalid close / OSBCLOSE.
The following IBM PTFs need to be applied that involves the IEC223I message.
UK02664
UK04264 (PRE UK01284)
UA18639
In one reported situation, simply applying PTF UK02664 eliminated the IEC223I message.
Important Note: this is not a TIBCO Rendezvous related problem; if you receive the IEC223I message, you should apply the above IBM PTFs on the mainframe machine.
The problem is a 'C' and LE run-time issue involving the function call fseek().
Additional information:
For the I/O code fixed by PQ96742 (PTF UK02664 for R709 = z/OS V1.6; and PTF UK02662 for R707 = z/OS V1.4), the module updated was EDCBP030 – this includes the sp30 code seen in the failing Traceback flow that contains:
“TIBCO” -> fseek -> sp30 -> osseek -> CEL4I031 -> osopenru -> OSBCLOSE
sp30 has code that maintains the amount of data in a data set, and it determines when a data set has been extended. In effect, it can determine when a multivol data set moved to another volume. This is what has happened here - the data set filled the first volume and moved to another volume. The code in sp30 could not handle this. sp30, upon determining that a data set had moved onto another volume invalidly *forced* a re-open of the data set (done by osopenru), so any new data that might have been written on the second volume was accounted for. This caused a close (OSBCLOSE) to be done as part of the re-open. The test to determine if a re-open is necessary, however, was bad. PQ96742 also fixed this test in sp30, so an invalid re-open is “not” done – and, the data on the new volume is simply part of the fseek (no re-open / OSBCLOSE necessary).
Therefore:
1) PQ96742 can lead to MSGIEC2231 and RC20 in certain environments
2) PQ96742 relates to data spanning volumes for a multi-volume data set. EDCBP030 was also fixed, so a re-open of a data set is not performed when volumes are crossed. This re-open was leading to an invalid close / OSBCLOSE.
Issue/Introduction
We are running RV on z/OS and saw the IEC223I message in the JOB LOG, what seems to be the problem?
Feedback
Yes
No
Powered by
