CollectingdiagnosisinformationforSAPHANA
(2022-12-06 21:57:33)
You can
-
In the Administration Editor in HANA Studio, choose
the
Diagnosis Files tab. If the system is offline, the Administration editor opens in diagnosis mode: -
-
Choose
Diagnosis Information > Collect -
Select whether
System Views should be included (in earlier HANA revisions this option was offered in the step before by choosing"Collect with SQL" or "Collect without SQL ) and thedate range for which you want to collect the diagnosis information. -
Click the "Finish Button"
to trigger the collection of diagnosis information -
See below section 'Where to Download' to download the
collection
-
Choose
-
You
are logged on as the operating system administrator adm on the command line: -
-
Change to
directory where the fullSystemInfoDump.py script is located (terminal shortcut "cdpy") i.e. $DIR_INSTANCE/exe/python_support -
For a database in a
single-container system please execute: python fullSystemInfoDump.py --days=
For a database in amultiple-container (MultiDb, MDC) system, you must also specify name of the tenant database with the corresponding option.
For a tenant database: python fullSystemInfoDump.py -t --days=
For the system database: python fullSystemInfoDump.py -t SYSTEMDB --days=
-
Change to
- You are logged on to DBACOCKPIT
-
-
Go to Diagnostics/Diagnosis Files, in the upper task bar click
"List Diagnosis
Information"
[Requires permissions "GRANT SELECT ON SYS.FULL_SYSTEM_INFO_DUMPS" (available in HANA Revision 101 and later), "EXECUTE on the procedure SYS.FULL_SYSTEM_INFO_DUMP", "EXECUTE on the procedure SYS.FULL_SYSTEM_INFO_DUMP_CREATE", "EXECUTE on the procedure SYS.FULL_SYSTEM_INFO_DUMP_RETRIEVE", "EXECUTE on the procedure SYS.FULL_SYSTEM_INFO_DUMP_DELETE"] -
- If systems views shall be included then the following permissions are required as well [SELECT on SYS_STATISTICS schema, SELECT on SYS_EPM schema, EXPORT privilege]
-
Go to
Diagnostics/Diagnosis Files, in the upper task bar click "Generate RTE Dump" [Requires permission "GRANT Resource Admin]
Those features require the following SAP_BASIS Support Packages: -
- 7.51 (all service packs)
- 7.50 (SP01 and higher)
- 7.40 (SP14 and higher)
- 7.31 (SP18 and higher)
- 7.30 (SP15 and higher)
-
7.02 (SP18 and higher)
-
Go to Diagnostics/Diagnosis Files, in the upper task bar click
"List Diagnosis
Information"
-
Open SAP HANA Cockpit
-
-
Click the tile "Resource Directory" - Select your SYSTEMDB entry
- Search for the tile "Alerting and Diagnostics", here click "Manage full system information dumps"
- In the right upper corner you find "Collect Diagnostics" -> "Collect from Existing Files"
- Keep the 7 day range if your issue was in this time or change appropriate.
- Then click "Start Collecting"
- HANA Cockpit will create a Full System Information Dump for the SystemDB and all Tenants.
-
Note:
The collection
represents
- The trace files of each host in the landscape i.e. $DIR_INSTANCE//trace/.
- The database configuration files.
-
The backup logs, crashdump files &
recorded performance trace files. -
A
runtime dump for each indexserver. - Among a set of other trace / configuration files, please refer to the documentation referenced for details
For storage reasons, by default only trace files from the last 7 days are collected unabridged (time range can be modified).
Where to Download
You can
- __.zip):
-
In the Administration Editor in HANA Studio, choose
the
Diagnosis Files tab, then select Diagnosis Information > Collect... to download the full-system-info-dump. -
Directly from the directory DIR_GLOBAL/sapcontrol/snapshots which typically points to /usr/sap//SYS/global.
As the size of the ZIP-archive created usually exceeds the maximum upload size of attachments in incidents, please either request a repository link or sFTP
(see KBA 2077673) share from the Support Engineer for uploading the trace files. Additional information to avoid running into known problems:
-
Do not collect the full-system-info-dump from
a standby host in distributed landscape when using SAP's Storage Connector (see SAP Note 2014225 for details). -
For HANA
revision 74.04 and lower please choose the command line option (see SAP Note 2040144 for details). -
In rare cases and before
HANA revision 80, the python call from the command line lead to high CPU usage. To limit the resources allocated by python, you can assign a lower priority to this process call by using "nice -19 python ...".
-
In the Administration Editor in HANA Studio, choose
the