Protokoll der 90. z/OS GUIDE-Tagung in Lahnstein
Transcription
Protokoll der 90. z/OS GUIDE-Tagung in Lahnstein
Protokoll der 90. z/OS GUIDE-Tagung in Lahnstein vom 04. bis 06. Oktober 2006 Allgemeines/Administratives Zu unserer diesjährigen Herbsttagung konnte Chairman Dietmar Frerix insgesamt 70 Teilnehmer und 9 Gäste in Lahnstein begrüßen. Wegen der geringen Teilnehmerzahl (und einer anderen Veranstaltung im Hause) tagten wir nicht im großen Sitzungssaal. Mitgliederbestand (alt) anwesende Teilnehmer entschuldigte Teilnehmer gelöschte Mitglieder (4x unentschuldigtes Fehlen) Abmeldungen/Ruhestand Neuanmeldungen Mitgliederbestand aktuell 153 70 39 4 4 3 148 Die Termine der nächsten zwei Tagungen: Frühjahrstagung Herbsttagung 07. – 09.03.2007 19. – 21.09.2007 Vorträge Die Vorträge der Tagung liegen auf dem neuen GSE Server http://gseimis2.gse.org unter Communities, SOSXD zOS (MVS) German Region, G#90 Oct2006 Die Vorträge von Herrn Schiefelbein und Herrn Benke Linux auf IBM System z – Update What’s new in RMF konnten wegen technischer Probleme (Firewall) nicht auf den GSE Server hochgeladen werden. Zusätzlich liegen alle Vorträge auf einem FTP Server beim KRZN http://guide.krzn.de Wir bedanken uns bei den Rednern – insbesondere den GSE Mitgliedern für die Erfahrungberichte - sowie bei Herrn Penzlin für die Verpflichtung der Referenten. Zum Vortrag CIM: mehr als nur Catalog-Recovery von Hostsystems als Nachtrag noch die email: [email protected] Requirements zwei Requirements wurden vorgestellt und verabschiedet mit Priority High File Latch Contention Up to now we have to provide a dump from USS address space including SYSZBPX2 data space to get information about the file latch contention. If a latch contention occurs we have to enter: ************************************************************** D GRS,LATCH,C ISG343I 15.28.43 GRS STATUS 066 LATCH SET NAME: SYS.BPX.A000.FSLIT.FILESYS.LSN.01 CREATOR JOBNAME: OMVS CREATOR ASID: 000E LATCH NUMBER: 10113 REQUESTOR ASID EXC/SHR OWN/WAI ************************************************************** In case of file system latch contention we have the possibility to use D OMVS commands to get useful information about file system name,owner, waiter and so on. For file latch contention, like shown in previous GRS command ouput, no commands are available to get additional hints. We do not know the file system nor the file name. This request is related to PMR 77041,070,724 von Frau Pfeiffer von der Filiadata GmbH in Karlsruhe WLM managed Batch im Sysplex beim Öffnen zusätzlicher Initators durch WLM soll – neben den goals die Auslastung der einzelnen Systeme berücksichtigt werden von Herrn Weilandt von der AOK Baden-Würtemberg in Stuttgart Berichte der Arbeitsgruppen zOS/390 (MVS) – Nord die Arbeitsgruppe trifft sich zweimal jährlich in Lüneburg (im Jahre 2006 im Juni und im Dezember), beim letzten Treffen standen folgende Punkte auf der Tagesordnung Erfahrungsbericht zum Health Checker neuer GSE Server gseimis2 IBM News Erfahrungsaustausch zOS/390 (MVS) – West die Arbeitsgruppe trifft sich jedes Jahr im Mai und im November bei unterschiedlichen Mitgliedern der Arbeitsgruppe, Tagesordnungspunkte beim letzten Treffen waren Java Batch neuer GSE Server gseimis2 IBM News Erfahrungsaustausch How to handle CBU and DR in a subcapacity environment Zu diesem Thema hat Herr Penzlin, IBM noch folgende Information Here are the guidelines we send out for CBU, Disaster Recovery and other "unusual" situations. Summary: Intervals with increased rolling 4-hour average utilization due to CBU Test or Disaster Recovery Test should not be considered when calculating SubCapacity software charging. IBM's Contractual Terms: CBU Upgrade & Disaster Recovery will not result in any increase or decrease in IBM Program charges to your Enterprise during a Test or Emergency. Customers their CBU Customers per year, using Capacity Back Up (CBU) must remain within the bounds of contract regarding the number of and duration of CBU tests. not using CBU should limit disaster recovery testing to 3 tests each test lasting no longer than 3 days. Capacity Back Up (CBU) and Disaster Recovery testing are considered “Unusual Situations” under IBMs SubCapacity policies and procedures. Any increases in MSU reporting attributable to CBU/DR test are not billable and can be overridden in SCRT reporting as long as the testing is within standard guidelines and conducted outside of peak periods. Unusual Situations Customers have the primary responsibility for preventing uncontrolled loops, operator errors, or unwanted utilization spikes. However, IBM understands that occasionally situations that could not be prevented (especially situations related to disaster recovery) may cause exceptional utilization values. In these situations, IBM does not expect our customers to pay for the increased utilization associated with the unusual situation. Please use your best judgement to determine if an unusual situation has occurred. IBM does not publish a list of unusual situations because they will by their nature be unpredictable. IBM trusts your judgement when you indicate an unusual situation has occurred. However, if you repeatedly override the Tool MSUs (for example, you always have unusual situations occur at quarter close) or if large discrepancies are found during IBM audits (IBM Call Home TSAD data is compared with data from subcapacity reports), IBM will contact you at the time the pattern is noticed. Modifying the SCRT Report to correct for "Unusual Situations" : If an unusual situation occurs, customers should NOT use IFASMFDP to delete SMF records from the SCRT report to be submitted to IBM for billing. Instead, the “Customer MSUs” fields, along with an explanation in “Customer Comments” should be used to submit the “correct” number of MSUs the customer has determined are associated with “normal” workload. Care should be taken to NOT change any values in the "Tool MSUs" column of the report. SCRT Override Measurement Process: Here are some step-by-step instructions on how to use SCRT to determine the override values: STEP 1) Run the SubCapacity Report against the entire month of data, including the days when the 'unusual event' occurred. THIS IS THE 'OFFICIAL' REPORT YOU WILL SUBMIT TO IBM. STEP 2) Check the Product Max Contributors Section of this SubCapacity Report to see if any of the product peaks occurred during the 'unusual event'. - If not, then your 'unusual circumstance' did not contribute to the billing-related peak utilization and no further action is needed. Submit the original report without overrides. - If so, proceed to step #3. STEP 3) Use the SMF Dump Program (IFASMFDP) to remove the SMF records for the time period during which the 'unusual event' occurred and create a new SMF data file. STEP 4) Run SCRT again, using the SMF subset dataset created in STEP 3. STEP 5) Compare the MSU values from the report generated in this step (Step #4) to those in the original report (STEP 1). Enter any changed values into the "Customer MSUs" column of the report you generated in STEP #1. Beside each MSU value, enter the comment "[Unusual Event Description] occurred from <date> to <date>". Submit both reports by the 9th day of the calendar month following data collection. Submit SCRT report generated in STEP 1 (and updated with overrides in STEP 5) using your normal submission process. Submit SCRT report generated in STEP 4 to " [email protected] " with a note saying "Backup report for [Unusual Event Description]" Keep ALL SMF70 and SMF89 records for 6months as prescribed by the SubCapacity Ts & C IBM Billing for each product: If the Data Collection % on the subset Report generated in STEP 4 is 90% or more: IBM will use the Override value from STEP 4. If the Data Collection % on the subset Report generated in STEP 4 is less than 90%: IBM will use the greater of: (1) the Override value from STEP 4 OR (2) the MSU values from the prior month. Please note: This procedure is subject to change at any time. Wolfgang Kleinen 23.10.2006