SAP Knowledge Base Article

2147247 - FAQ: SAP HANA Statistics Server

Component: HAN-DB-MON (SAP HANA > SAP HANA Database > SAP HANA Monitoring), Version: 207,
Released On: 25.03.2025

Symptom
You are interested in the SAP HANA statistics server.

Environment
SAP HANA

Cause
1. What is the SAP HANA statistics server?
2. Where do I find more information about the SAP HANA statistics server?
3. Which indications exist for problems in the area of the SAP HANA statistics server?
4. How is the statistics server implemented?
5. How can I determine details about the statistics server?
6. What are the advantages of the embedded statistics server?
7. How can the standalone statistics server be migrated to the embedded statistics server?
8. How can the statistics server be configured?
9. How can the memory requirements of the statistics server be minimized?
10. How can the runtime and CPU requirements of the statistics server actions be analyzed and optimized?
11. Why is the history collection for some tables inactive?
12. Why are some statistics server actions disabled?
13. How can I take care for disabled statistics server actions?
14. How can I check for problems in the area of the statistics server?
15. What is a good retention time for statistics server histories?
16. What can I do if a statistics server action fails with "301: unique constraint violated"?
17. Is it possible to mark an alert as acknowledged or handled?
18. Are the default alert thresholds generally sufficient?
19. How can the statistics server check intervals be adjusted?
20. How can internal statistics server errors be analyzed?
21. What happens if the STATISTICS_SCHEDULE table is manually modified in a wrong way?
22. How can I analyze and resolve SAP HANA alerts?
23. Is it possible to restart the statistics server?
24. What is the difference of SNAPSHOT_ID and SERVER_TIMESTAMP in histories?
25. Why do different statistics server actions run in the system database and tenants?
26. How can I monitor and tackle issues with event acknowledgement?
27. How can I check the state of the statistics server and restart it if required?
28. Can Native Storage Extension (NSE) be used in context of the statistics server tables?
29. Are there SQLScript version requirements for the statistics server?
30. Do statistics server histories also contain information for remote system replication sites?

Resolution

1. What is the SAP HANA statistics server?

The SAP HANA statistics server (or: statistics service) is a central tool for monitoring the SAP HANA database. Among others
it performs the following tasks:
Regular check for critical situations and generation of alerts
Regular history collection of monitoring information into histories located in the _SYS_STATISTICS schema.
2. Where do I find more information about the SAP HANA statistics server?
See section "The Statistics Service" in the SAP HANA Administration Guide for more information about the statistics server.
The definition of views used by the embedded statistics server can be found in the Embedded Statistics Server Views
Reference.

3. Which indications exist for problems in the area of the SAP HANA statistics server?
The following SAP HANA alerts indicate problems in the SAP HANA statistics server area:

Alert Name SAP Note Description

0 Internal statistics server problem 2081609 Identifies internal statistics server problem.

SQL: "HANA_Configuration_MiniChecks" (SAP Notes 1969700, 1999993) returns a potentially critical issue (C = 'X') for one
of the following individual checks:

Check ID Details

M0710 Open alerts (high priority)

M0711 Open alerts (error state)

M0712 Internal statistics server errors (last day)

M0715 Number of actions not executed as expected

M0716 Number of statistics server worker threads

M0717 Number of disabled actions

M0718 Number of relevant inactive actions

M0719 Number of actions with unknown state

M0720 Events not acknowledged since >= 1800 s

M0721 Thread call stack collector inactive

M0730 Pending e-mails older than 3 days

M0735 Alerts older than 42 days

M0736 Alerts reported frequently

M0740 Time since statistics server run (s)

M0744 Total SQL cache share of statistics server (%)

M0745 Total size of statistics server tables (GB)

M0746 Total memory share of statistics server (%)

M0747 Number of zero entries in HOST_SQL_PLAN_CACHE

M0748 History of M_CS_UNLOADS collected

M0749 History of M_RECORD_LOCKS collected

M0750 Stat. server tables with retention < 42 days

M0751 Historic thread samples save interval (s)

M0752 Historic thread call stacks interval (s)

M0753 Retention time of table disk sizes (days)

M0754 Historic long running statements interval (s)

 M0755 Embedded statistics server used

M0760 Status of embedded statistics server migration

M0762 Executed DDL statements (short-term)

M0764 Avg. manual stat. server sessions (short-term)

M0765 Log segment size of statisticsserver (MB)

M0766 Workload class for statistics server in place

M0770 Number of stat. server tables not on master

M0780 Unknown entries in HOST_OBJECT_LOCK_STATISTICS

M0781 Orphan entry for alert 32

4. How is the statistics server implemented?

The statistics server is either implemented embedded (ESS) or standalone (SSS):

Statistics server t Shortc Default Recommen Details

ype ut ded

Standalone statistic SSS Rev. <= 1.0 Rev. <= 1.00 Old approach using a dedicated process (statisticsserver), various disa
s server 0.92 .73 dvantages:

Increased memory requirements

Unnecessary large history data collections (e.g. HOST_SQL_PL
AN_CACHE)
Important history data not collected (e.g. HOST_SERVICE_TH
READ_SAMPLES)
No shared resources with indexserver process, so increased risk
of OOM and instabilities

Embedded statistics ESS Rev. >= 1.0 Rev. >= 1.00 New approach embedded into indexserver process
server 0.93 .74

SQL: "HANA_TraceFiles_MiniChecks" (SAP Notes 1969700, 2380176) reports one of the following check IDs:

Check ID Details

T0101 Unique constraint violation

T0110 Insert into HOST_SERVICE_THREAD_SAMPLES failed

T0112 Error sending e-mail

5. How can I determine details about the statistics server?

SAP Note 1969700 provides the following SQL statements for retrieving details about the statistics server:

SQL statement Statistics se Details

rver type

SQL: "HANA_StatisticsServer_Ale all Overview of currently reported alerts

rts_Current"

SQL: "HANA_StatisticsServer_Ale all Available alerts including description and recommendation

rts_Definition"

SQL: "HANA_StatisticsServer_Ale all Overview of alerts reported in history

rts_History"
 SQL: "HANA_StatisticsServer_His all Retention time and actual data of statistics server histories
tories_RetentionTime"

SQL: "HANA_StatisticsServer_Mai ESS Overview of existing e-mail notifications, amount can increase considerably if
lProcessing" e-mail is configured improperly, see SAP Note 2133799

SQL: "HANA_StatisticsServer_Pri all Checks for primary keys are defined on statistics server histories, can be remo
maryKeysOnHistoryTables" ved in exceptional situations as described in SAP Note 2143679

SQL: "HANA_StatisticsServer_Pro ESS General properties, can be used to check if migration to ESS was successful
perties"

SQL: "HANA_StatisticsServer_Sch ESS Schedule of statistics server actions (e.g. frequency, last run, runtime per hour
edule" )

Statistics server alerts can be monitored and configured in SAP HANA Studio -> Administration -> Alerts.

6. What are the advantages of the embedded statistics server?

The main advantages of the embedded statistics server are:
Reduced memory footprint due to more efficient data processing
More comprehensive history collections (e.g. HOST_SERVICE_THREAD_SAMPLES, see SAP Note 2114710)
Elimination of instabilities and stuck situations caused by indexserver / statisticsserver process separation
Updated alert checks
Error-prone manual adjustments of parameters (see e.g. SAP Note 2084747) no longer required
Prerequisite for specific SAP HANA features (e.g. keeping row store in memory during restart, see SAP Note 2159435)
Extended retention time of 42 days for most tables
Standalone statistics server is no longer maintained by SAP

7. How can the standalone statistics server be migrated to the embedded statistics server?
The standalone statistics server is typically active when the following parameter is set:

statisticsserver.ini -> [statisticsserver_general] -> active = true

A manually triggered migration to the embedded statistics server can be performed based on SAP Notes 2092033 and
1917938. See also SAP Note 1925684 that describes prerequisites from SAP ABAP and Solution Manager perspective that
need to be fulfilled.
The main technical activation for the embedded statistics server is the following parameter setting:
System DB, standalone DB:

nameserver.ini -> [statisticsserver] -> active = true

Tenant DB:

indexserver.ini -> [statisticsserver] -> active = true

When installing or upgrading to SAP HANA 1.00.93 or higher, the embedded statistics server is activated automatically.
If the embedded statistics server disables itself automatically or doesn't collect data, see SAP Notes 2006652,
2036630 and 2091256 for corrective measures. Additionally you can proceed according to the troubleshooting tree provided
in SAP Note 2223237. In case of corruptions in statistics server objects you can also consider SAP Note 2343366 for repair /
re-creation steps.

8. How can the statistics server be configured?

Some alert settings can be configured in SAP HANA Studio (-> "Administration" -> "Alerts"). For more specific
configurations the details depend on the type of the statistics server.
Standalone statistics server:
The configuration is done via SAP HANA parameters and should be handled with care. SAP Notes 2084747 and 1929538 for
some examples.
Embedded statistics server:
The configuration is done based on changes of configuration tables. See SAP Notes 1991615 and 2529478 for details.

9. How can the memory requirements of the statistics server be minimized?

The statistics server may collect much more data than required. In order to minimize its memory footprint you
should perform the following actions based on the type of the statistics server. All proposed changes have no impact on the
analysis capabilities, but they can help to reduce the memory footprint significantly.
General actions:
Avoid retention times of more than 42 days in order to limit the size of the histories. You can check the currently
configured retention times and the actual data retention via SQL: "HANA_StatisticsServer_Histories_RetentionTime"
(SAP Note 1969700). Exception: Historic table disk sizes (ID 5010) should be retained for a significantly longer time as
suggested in check ID M0753 ("Retention time of table disk sizes (days)") of the SAP HANA Mini Checks (SAP Note
1999993).
Standalone statistics server:
If you are at least on SAP HANA 1.00.74 it is recommended to switch to the embedded statistics server which will no
longer use these allocators (see SAP Notes 1917938 and 2092033 for more information).
If you are already in an OOM situation of the statisticsserver it might be difficult to perform the cleanup operations
suggested below, because they fail due to the OOM situation (e.g. "table update failed"). In this case you can temporarily
increase the SAP HANA parameter statisticsserver.ini -> [memorymanager] -> allocationlimit (e.g. to 10% or 20%) and
perform the cleanup activities. Afterwards you should unset the allocationlimit parameter because the default should be
sufficient when the below optimizations are implemented.
Add "WHERE EXECUTION_COUNT > 0" to the extraction command for HOST_SQL_PLAN_CACHE as described in
SAP Note 2084747.
Manually delete existing records with EXECUTION_COUNT = 0 from HOST_SQL_PLAN_CACHE:
DELETE FROM _SYS_STATISTICS.HOST_SQL_PLAN_CACHE WHERE EXECUTION_COUNT = 0

Disable the history collection of M_CS_UNLOADS in HOST_CS_UNLOADS as described in SAP Note 2084747.
Check according to SAP Note 2084747 if there are other unnecessarily large histories and adjust the data collection if
required.
Embedded statistics server:
Manually delete existing records with EXECUTION_COUNT = 0 from HOST_SQL_PLAN_CACHE:
DELETE FROM _SYS_STATISTICS.HOST_SQL_PLAN_CACHE_BASE WHERE EXECUTION_COUNT = 0

If you are below Revision 85, you should perform this DELETE operation on a regular basis (e.g. weekly).
Once you have upgraded to Revision 85 or higher, the embedded statistics server will massively reduce the collected data
of HOST_SQL_PLAN_CACHE and no manual intervention is required.
Disable the data collection of M_CS_UNLOADS in HOST_CS_UNLOADS (if it is still active):
UPDATE _SYS_STATISTICS.STATISTICS_SCHEDULE SET STATUS = 'Inactive' WHERE ID = 5035;
TRUNCATE TABLE _SYS_STATISTICS.HOST_CS_UNLOADS_BASE

Disable the data collection of M_RECORD_LOCKS in HOST_RECORD_LOCKS (if it is still active):

UPDATE _SYS_STATISTICS.STATISTICS_SCHEDULE SET STATUS = 'Inactive' WHERE ID = 5033;
TRUNCATE TABLE _SYS_STATISTICS.HOST_RECORD_LOCKS_BASE

Check "How can I monitor and tackle issues with event acknowledgement?" in order to rule out an increased amount of
non-acknowledged events.
Make sure that there is no high amount of records for single alerts in table STATISTICS_ALERTS_BASE, e.g. via SQL:
"HANA_StatisticsServerAlerts_FrequentAlerts_ESS" (RETAINED_RECORDS_PER_ALERT = 500000) available via
SAP Note 1969700. Particularly high number of records can exist due to a bug with alert 65 that is fixed with Rev.
1.00.97.02 and 1.00.101.
On SAP HANA SPS 09 and below you can implement SAP Note 2186299 in order to make sure that the size of heap
allocator Pool/RowEngine/MonitorView doesn't grow unnecessarily.
A large size of HOST_LOAD_HISTORY_SERVICE and HOST_LOAD_HISTORY_HOST can be caused by a small
setting of parameter global.ini -> [resource_tracking] -> load_monitor_granularity. Make sure that at least a value of
1000 ms is used for that setting. See SAP Note 2222110 for more information.
If HOST_OBJECT_LOCK_STATISTICS contains a very large number of entries, you can increase the collection interval
of the related ID 5032, e.g. from one minute to 10 minutes. Starting with Rev. 1.00.102.01 the data collection is changed
 so that only long running lock waits (>= 1 s) are historicized instead of all.
The transactional lock history in HOST_OBJECT_LOCK_STATISTICS may contain a high amount of entries for
OBJECT_NAME = '(unknown)'. They typically refer to already dropped temporary tables and can be deleted. This
problem is fixed starting with SAP HANA 2.0 SPS 00. You can consider using SAP HANACleaner (SAP Note 2399996)
to configure an automatic cleanup.
DELETE FROM _SYS_STATISTICS.HOST_OBJECT_LOCK_STATISTICS_BASE WHERE OBJECT_NAME = '(unknown)'

See SAP Note 2170779 and make sure that table old records are deleted from table STATISTICS_ALERTS_BASE, e.g. by
executing the following command. Starting with SAP HANA 1.0 SPS 11 the statistics server will automatically purge old
alerts after 42 days:
DELETE FROM _SYS_STATISTICS.STATISTICS_ALERTS_BASE WHERE ALERT_TIMESTAMP < ADD_DAYS(CURRENT_TI
MESTAMP, -42)

10. How can the runtime and CPU requirements of the statistics server actions be analyzed and optimized?
All checks and collections triggered by the statistics server are executed with the following wrapper call (statement hash
d6fd6678833f9a2e25e7b53239c50e9a):

call _SYS_STATISTICS.STATISTICS_SCHEDULABLEWRAPPER('Timer', ?, ?, ?, ?)

You can check its load key figures by executing SQL: "HANA_SQL_StatementHash_KeyFigures" (STATEMENT_HASH =
'd6fd6678833f9a2e25e7b53239c50e9a') available via SAP Note 1969700. If you observe a significant average runtime (e.g.
time per execution >= 500 ms), you can have a closer look at the individual actions by executing SQL:
"HANA_StatisticsServer_Schedule" (ORDER_BY = 'TIME_PER_H') available via SAP Note 1969700. The following output
illustrates a case where the collector for the HOST_BLOCKED_TRANSACTIONS histories is responsible for significant
execution times of more than 600 seconds per hour:

You can now find a good trade-off between the load and the collected data. If it is e.g. okay that the history of blocked
transactions is only collected every five minutes instead of every single minute, you can increase the interval from 60 to 300
seconds:

UPDATE _SYS_STATISTICS.STATISTICS_SCHEDULE SET INTERVALLENGTH = 300 WHERE ID = 5020

For the following alerts and collectors special considerations are useful if they are responsible for a high runtime or resource
consumption:

I Action Details
D

65 Alert_Bac The runtime of backup related actions can be impacted by a large backup catalog. See check ID M0940 (SAP Note 1
kup_Long 999993) and SAP Note 2505218 and make sure that the catalog remains on a reasonable size.
_Log_Bac
kup

41 Alert_Dec The Alert_Dec_Extractor_Status and Collector_Global_Dec_Extractor_Status checks don't work correctly and ha
50 _Extractor ve an increased runtime as long as the CATALOG READ privilege isn't assigned to the _SYS_STATISTICS user. Th
0 _Status erefore you should either grant CATALOG READ to _SYS_STATISTICS or disable the check with ID 5008. See SAP
8 Collector_ Note 2000002 ("Are there standard recommendations for specific SQL statements available?" -> "ba8161b6133633
Global_De e2984828ce38aa669a, 905dbaa93a672b087c6f226bc283431d") for more details.
c_Extract
or_Status

50 Collector_ This collector can suffer from the existence of large row store tables. See SAP Note 2000002 ("Are there standard r
01 Global_Ro ecommendations for specific SQL statements available?" -> "0c2d40de2702bc576de87a9aec2cef30, f45e2eaba236
wstore_Ta b6655804f6585d631923") for more details.
bles_Size
 50 Collector_ For this collector, an execution frequency of 5 minutes is sufficient (issue number 334265). See SAP Note 2000002
2 Host_Lon ("Are there standard recommendations for specific SQL statements available?" -> COLLECTOR_HOST_LONG_R
8 g_Runnin UNNING_STATEMENTS) for more details.
g_Stateme
nts

50 Collector_ This collector can suffer from the existence of large row store indexes. See SAP Note 2000002 ("Are there standard
45 Host_RS_ recommendations for specific SQL statements available?" -> "ac52398f58a752bed7843aac8d829464, e9f2feac54a0
Indexes 48984928f3fcbcfacedd") for more details.

50 Collector_ The runtime of this collector can be increased if the secondary system replication site is not accessible. See SAP Not
46 Host_Serv e 2000002 ("Are there standard recommendations for specific SQL statements available?" -> "M_SERVICE_REPL
ice_Replic ICATION") for more details.
ation

The maximum parallelism of the statistics server threads is controlled by the following parameter (default: 5):

<service>.ini -> [statisticsserver] -> threadpool

It is usually not recommended to adjust this parameter, instead workload classes as suggested below should be used.
For SAP HANA <= 2.0 SPS 05 see SAP Note 2970921 and make sure that a workload class is defined for the
_SYS_STATISTICS user in order to avoid temporary CPU spikes.
With SAP HANA >= 2.0 SPS 06 an Embedded Statistics Service Workload Class is automatically implemented in case no
workload class for the _SYS_STATISTICS user exists, yet. When the delivered workload class is used, the following parameter
is set to its default 'yes':

indexserver.ini -> [statisticsserver] -> enable_automatic_workload_management

In case a workload class already exists from the past, the parameter is set to 'no'. If you want to enable the delivered workload
class, you need to drop the previous workload class and subsequently unset the parameter so that its default 'yes' becomes
active. See SAP Note 2222250 for more information related to workload classes.
Be aware that the suggested workload class is based on the database user name ('USER NAME' = '_SYS_STATISTICS'). In the
thread information (SAP Note 2114710) the database user name may be empty for statistics server activities and only the
application user name is filled with _SYS_STATISTICS. This topic is tracked via issue number 266415 and has no impact on
the validity of the defined workload class.
Also check "How can the memory requirements of the statistics server be minimized?" in terms of
STATISTICS_ALERTS_BASE recommendations, because a high number of records in this table can also have CPU and
runtime effects.
The following wrapper call is typically triggered by Solution Manager monitoring (statement hash
dc571bf5eb7cad9c313c20de904ab709):

CALL _SYS_STATISTICS.STATISTICS_SCHEDULABLEWRAPPER( ? , ? , ? , ? , NULL)

If SAP HANA >= 1.00.122.04 is in place, you can activate the new SAP HANA monitoring mechanism of the Solution
Manager (SAP Note 2374272). Then the Solution Manager will rely on information already collected by the statistics server
rather than executing statistics server checks on its own. As a consequence this particular
STATISTICS_SCHEDULABLEWRAPPER call will no longer be executed.

11. Why is the history collection for some tables inactive?

SQL: "HANA_Configuration_MiniChecks" (SAP Note 1969700) may report an issue for check ID M0718 ("Number of
relevant inactive actions"), because some history collections are deactivated with certain release levels of the embedded
statistics server, although it is normally useful to collect this information.
The following checks are set to 'Inactive' on purpose and don’t need to be activated:
Alert_Check_Free_Physical_Mem (ID 1)
Alert_Column_Tables_Size_Growth (ID 20)
Alert_Dec_Extractor_Status (ID 41)
Alert_Plan_Cache_Size (ID 58)
Alert_Row_Store_Version_Space_Overflow (ID 73)
 Alert_Metadata_Version_Space_Overflow (ID 74)
Alert_Check_Database_Disk_Usage (ID 77)
Alert_Check_DQ_Reference_Data_Expiring (ID 95)
Alert_Check_DQ_Long_Running_Task_Plan (ID 96)
Alert_Mon_Column_Tables_Record_Count_Incl (ID 117)
Alert_Partitioned_Table_Record_Count_Incl (ID 127)
Collector_Global_Dec_Extractor_Status (ID 5008)
Collector_Host_Connections (ID 5024)
Collector_Host_Connection_Statistics (ID 5025)
Collector_Host_Record_Locks (ID 5033)
Collector_Host_CS_Unloads (ID 5035)
Collector_Host_CS_Indexes (ID 5059)
Be aware that there can be good reasons to individually deactivate checks, e.g. when the check functionality (like LOB
reclaims or a consistency check) is triggered in a different way or when bugs need to be bypassed.
In order to activate relevant history collections you can run the following command (usually not required in system
databases) after having adjusted <list_of_individually_deactivated_checks>:

UPDATE "_SYS_STATISTICS"."STATISTICS_SCHEDULE" SET STATUS = 'Idle' WHERE STATUS = 'Inactive' AND I

D NOT IN (1, 20, 41, 58, 73, 74, 77, 95, 96, 117, 127, 5008, 5024, 5025, 5033, 5035, 5059[, <list_
of_individually deactivated_checks>])

Attention: Be aware that in system databases of MDC environments (SAP Note 2101244) several other actions are
deactivated per default starting with SAP HANA 2.0 SPS 04. This is normal and there is usually no need for adjustments.

12. Why are some statistics server actions disabled?

Depending on the type and release of the statistics server alerts and history collections are disabled in the following scenarios:

Ty Revision Details
pe

SS general The standalone statistics sever never disables individual actions.

S

ES 1.0 SPS 0 Individual actions are disabled if the previous execution is not finished when the next execution is started.
S 7
Disabled actions won't be enabled automatically, you can only enable them manually.
1.0 SPS 0
8 <= Rev.
85.03
1.0 SPS 0
9 <= Rev.
92

ES 1.0 SPS 0 Individual actions are disabled if the previous execution is not finished when the next execution is started.
S 8 >= Rev.
Disabled actions are skipped at least once and at least for one hour. Afterwards they are automatically enabled by t
85.04
he statistics server. So roughly speaking actions with a high execution frequency are enabled after one hour and acti
1.0 Rev. >
ons with a low execution frequency are enabled after they have been skipped once.
= 93

13. How can I take care for disabled statistics server actions?
Disabled statistics server actions are particularly critical with ESS and Revisions <= 1.00.92, because no automatic enabling
happens.
Disabled statistics server actions are reported with check ID M0717 ("Number of disabled actions") of SQL:
"HANA_Configuration_MiniChecks" (SAP Note 1969700).
The most frequent root cause for disabled statistics server actions are timeouts. You can check via SQL:
"HANA_StatisticsServer_Schedule" (SAP Note 1969700) for STATUS, which contains "Disabled" and the related reason,
e.g. "Disabled (timeout)".
If you find disabled actions due to timeout, you can enable them with the following command:
 UPDATE "_SYS_STATISTICS"."STATISTICS_SCHEDULE" SET STATUS = 'Idle' WHERE STATUS = 'Disabled' AND S
TATUSREASON = 'timeout'

If the same check disables itself repeatedly due to timeouts, you have to analyze the underlying performance problem. A
timeout is triggered, if the previous execution is still active when the next execution is triggered, so as a workaround you can
reduce the risk of timeouts by reducing the execution frequency of the involved action:

UPDATE _SYS_STATISTICS.STATISTICS_SCHEDULE SET INTERVALLENGTH = <interval_s> WHERE ID = <action_id

>

Another reason for disabled statistics server actions can be invalid underlying procedures. A general activation of actions
regardless of the timeout reason can be performed with the following command:

UPDATE "_SYS_STATISTICS"."STATISTICS_SCHEDULE" SET STATUS = 'Idle' WHERE STATUS = 'Disabled'

See SAP Note 2113228 for more information.

14. How can I check for problems in the area of the statistics server?
You can run SQL: "HANA_Configuration_MiniChecks" (SAP Notes 1999993 and 1969700) and have a look at the check
results for the "STATISTICS SERVER" section starting with CHECK_ID 700. If potentially critical issues are reported (C =
'X'), you can perform an analysis as described in SAP Note 1999993 and SAP Notes referenced there, and take appropriate
actions if required.

15. What is a good retention time for statistics server histories?

Experience has shown that a retention time of 6 weeks (42 days) is a good compromise between size and analysis capabilities.
Having 6 weeks of history available it is always possible to compare current issues with a time of the same load one month
back (e.g. month's end closing).
The retention time for most of the statistics server histories is 42 days on tenant level and 21 days on system DB level. Check
ID M0750 ("Stat. server tables with retention < 42 days") of SQL: "HANA_Configuration_MiniChecks" (SAP Note 1969700)
reports tables where the retention time is too small.
For the following collectors a reduced retention time is useful or acceptable:
5008 (Collector_Global_Dec_Extractor_Status)
5024 (Collector_Host_Connections)
5025 (Collector_Host_Connection_Statistics)
5026 (Collector_Global_Internal_Events)
5033 (Collector_Host_Record_Locks)
5035 (Collector_Host_CS_Unloads)
5600 (Collector_Streaming_Projects_Statistics)
5601 (Collector_Streaming_Publishers_Statistics)
5602 (Collector_Streaming_Subscribers_Statistics)
5603 (Collector_Streaming_Streams_Statistics)
5604 (Collector_Streaming_Project_Connections_Statistics)
For all others you can adjust the retention time to 42 days with the following command:

UPDATE _SYS_STATISTICS.STATISTICS_SCHEDULE SET RETENTION_DAYS_CURRENT = 42

WHERE RETENTION_DAYS_CURRENT < 42 AND ID NOT IN ( 5008, 5010, 5024, 5025, 5026, 5033, 5035, 5600,
5601, 5602, 5603, 5604 ) AND ID >= 1000

Having a higher retention value (e.g. 365 days) for certain histories is usually possible without problems. You can take the
original table size with 42 days history and estimate the expected target table size for the new history retention (e.g. about
factor 9 for 365 days). Sizes of up to a few GB can still be justified.

16. What can I do if a statistics server action fails with "301: unique constraint violated"?
As a workaround the primary keys of the involved histories can be dropped as described in SAP Note 2143679.
Duplicate keys are typically related to statistics server bugs. The following scenarios are currently known:
 Tables SAP Not Details Fixe
e / issue d wi
number th

GLOBAL_OUT_OF Collector 5053 (Collector_Global_Out_Of_Memory_Events) can suffer from unique const 2.00
_MEMORY_EVEN raint related terminations. .037.
TS 02
2.00
.041

COLLECTOR_GLO Row store tables with several containers can result in duplicate entries in M_RS_TABLES 2.00
BAL_ROWSTORE with SAP HANA 2.00.040 - 2.00.042. As a workaround you can reorganize these tables, e. .043
_TABLES_SIZE_B g. using SQL: "HANA_Tables_RowStore_TablesWithMultipleContainers" (SAP Note 196
ASE 9700).

COLLECTOR_HOS 235390 The primary key of the underlying table HOST_CS_INDEXES_BASE was designed withou 2.00
T_CS_INDEXES t SCHEMA_NAME and without TABLE_NAME / PART_ID. Therefore duplicates are poss .047
ible when indexes of the same name exist in more than one schema or when indexes are de
fined on partitioned tables.

COLLECTOR_HOS 259404 The collector is based on view M_CS_INDEXES_FOR_STATISTICSSERVER_ that lacks a in pr

T_CS_INDEXES PART_ID join condition. As a consequence duplicate keys are possible in context of partiti ogre
oned tables. ss

COLLECTOR_TEL The parameter telemetry collector can terminate with unique constraint violation when par 2.00
_INIFILE_CONTE ameters are maintained on HOST level (SAP Note 2186744). .040
NTS

HOST_JOB_HIST 2347981 1.00.

ORY_BASE 122

HOST_LONG_RU 1.00.
NNING_STATEME 122.
NTS_BASE 20
2.00
.023

HOST_SERVICE_ 2137142 Several threads with THREAD_ID -1 may exist in M_SERVICE_THREAD_SAMPLES. 1.00.
THREAD_SAMPL 90
ES_BASE

HOST_SQL_PLAN 2.00
_CACHE_OVERVI .024
EW_BASE .08
2.00
.036

HOST_VOLUME_I 2127852 1.00.

O_DETAILED_ST 112.
ATISTICS_BASE 02
HOST_VOLUME_I
O_RETRY_STATIS
TICS_BASE
HOST_VOLUME_I
O_TOTAL_STATIS
TICS_BASE

STATISTICS_ALE The primary key doesn't contain table partitions. If the same alert is issued for different pa 1.00.
RTS_BASE rtitions of the same table, the "unique constraint violated" error is issued. 110
 STATISTICS_ALE In context of alert ID 39 (long running statements) the unique index is defined on SNAPS 1.00.
RTS_BASE HOT_ID, ALERT_ID and TRANSACTION_ID. This is not generally sufficient in scale-out 122.1
systems, because the same transaction ID can exist on different hosts. 5
2.00
.012.
04
2.00
.023

STATISTICS_ALE In context of alert ID 49 (blocked transactions) duplicate keys are possible if different upda 1.00.
RTS_BASE te transaction IDs are blocked by the same blocker. 122.1
5

STATISTICS_ALE In context of alert ID 49 (blocked transactions) duplicate keys are possible if the same con 2.00
RTS_BASE nection ID exists on different scale-out nodes. .047

STATISTICS_ALE 259567 In context of alert ID 49 (blocked transactions) duplicate keys are possible due to inadequa 2.00
RTS_BASE te join operations. .048
.06
2.00
.056

STATISTICS_ALE 248283 Alert ID 75 (Alert_Row_Store_Version_Space_Skew) can also suffer from "unique constra 2.00
RTS_BASE int violated" errors. .048
.04
2.00
.055

STATISTICS_ALE 253739 In context of alert ID 94 (Alert_Replication_Logreplay_Backlog) duplicate keys are possib 2.00
RTS_BASE le if multiple secondary sites of the same primary site suffer from the backlog .048
.03
2.00
.053

STATISTICS_ALE 312954 Alert ID 131 (Alert_Admission_Control_Enqueue_Event) can suffer from duplicate keys d 2.00
RTS_BASE ue to a missing join condition in the procedure definition. .059.
15
2.00
.079.
01

17. Is it possible to mark an alert as acknowledged or handled?

Currently there is no option to mark an alert as acknowledged or handled, so you may still see alerts that you have already
handled and you may still get new alerts for topics you have already acknowledged and you are working on it. SAP is working
on an implementation of this feature.

18. Are the default alert thresholds generally sufficient?

The alert thresholds can be checked via SQL: "HANA_StatisticsServer_Alerts_Definition" (SAP Note 1969700) and should
be considered as initial proposal, but not as a general best practice. This means, you can adjust thresholds based on your
needs. Typical adjustments are:

Ale Alert name Details

rt I
D

17 Record count of Alerts may be issued quite early recommending partitioning of tables. Taking into account the risk of side e
non-partitioned c ffects of partitioning, it can be useful to increase the thresholds and implement / extend partitioning only
olumn-store tabl when the number of records gets closer to the 2 billion record limit. See SAP Note 2044468 for more infor
es mation related to partitioning.
 27 Record count of c Alerts may be issued quite early recommending partitioning of tables. Taking into account the risk of side e
olumn-store tabl ffects of partitioning, it can be useful to increase the thresholds and implement / extend partitioning only
e partitions when the number of records gets closer to the 2 billion record limit. See SAP Note 2044468 for more infor
mation related to partitioning.

50 Number of diagn Up to SAP HANA 1.0 SPS 08 the number of trace files is counted on a system-wide basis and not per host.
osis files This can result in misleading alerts. As a workaround you can increase the threshold for the number of trac
e files. Starting with 1.0 SPS 09 this adjustment is normally no longer required.

77 Database disk us The default thresholds (e.g. 300 / 400 / 500 GB) may not match with your system size and so you can adju
age st them based on your needs.

Be aware that in SAP HANA Studio you can only adjust thresholds for the "low", "medium" and "high" level. Thresholds for
the "info" level can be adjusted with SAP HANA Cockpit (SAP Note 2800006).

19. How can the statistics server check intervals be adjusted?

Adjusting the interval of a specific alert or collector can be done by updating the STATISTICS_SCHEDULE table:

UPDATE _SYS_STATISTICS.STATISTICS_SCHEDULE SET INTERVALLENGTH = <length_in_s> WHERE ID = <id>

If you e.g. want to make sure that the history of blocked transactions (ID = 5020) is only collected every five minutes instead
of every single minute, you can increase the interval from 60 to 300 seconds:

UPDATE _SYS_STATISTICS.STATISTICS_SCHEDULE SET INTERVALLENGTH = 300 WHERE ID = 5020

It is generally recommended that the history of thread samples (ID = 5034) is only collected at least every 10 minutes in order
to avoid too much data loss in case of a crash or shutdown (check ID M0751). If the current interval is higher you can reduce
it to 10 minutes (600 seconds) with the following command:

UPDATE _SYS_STATISTICS.STATISTICS_SCHEDULE SET INTERVALLENGTH = 600 WHERE ID = 5034

20. How can internal statistics server errors be analyzed?

Internal statistics server errors are reported with alert ID 0. You can display them by running SQL:
"HANA_StatisticsServer_Alerts_History" (ALERT_ID = 0) available via SAP Note 1969700.
Internal statistics server errors can result in a red icon in "Check for disabled collectors" in transaction DBACOCKPIT and in
the detailed screen of this check you see "Collectors are disabled: <collector>".
The following table contains typical internal statistics server errors and resolution steps:

E Error text Ac Details

r tio
r nI
o Ds
r

7 feature not supported: invalid charact 50 Due to the SAP HANA bug described in SAP Notes 2453766 and 2478767 certain
er encoding 34 column values of M_SERVICE_THREAD_SAMPLES can't be processed correctly
when being historicized to HOST_SERVICE_THREAD_SAMPLES (SAP HANA <
= 1.00.122.11, <= 2.00.012.00, 2.00.020). As a workaround the statistics server pr
ocedure can be adjusted based on the attachment to these SAP Notes.

2 invalid table name 60 This is a bug with SAP HANA 1.00.102.03 that is caused by the syntax change des
5 0 cribed in SAP Note 2241598. You can correct the procedures based on SAP Note 2
9 60 259611.
1

2 invalid column name 50 This is a bug with SAP HANA 1.00.102 to 1.00.102.01, see SAP Note 2239843 for
6 29 more information.
0
2 inserted value too large for column 50 This error can happen with SAP HANA 2.00.000 to 2.00.011 and it is typically lin
7 09 ked to an long FILESYSTEM_TYPE value caused by the fact that a file system typ
4 e is marked as unsupported (e.g. 'UNSUPPORTED (ext4)'). See SAP Note 199993
0 ("Are all types of filesystems supported for SAP HANA?") and make sure that on
ly supported file systems are in place.

2 inserted value too large for column 57 This termination can happen with SAP HANA 2.00.070 - 2.00.071 in context of C
7 05 ollector_Tel_Out_Of_Memory_Events when the string TOTAL_STATEMENT_
4 MEMORY_LIMIT_FROM_WORKLOAD_CLASS is inserted into column EVENT
_REASON, because it exceeds the defined limit of 32 characters (issue number 30
8466). You can use the normal OOM history HOST_OUT_OF_MEMORY_EVEN
TS collected by Collector_Global_Out_Of_Memory_Events that already uses a su
fficient EVENT_REASON length limit of 64 characters.

Analysis commands available via SAP Note 1969700 generally rely on the properl
y working history HOST_OUT_OF_MEMORY_EVENTS, so there is no risk of mi
ssing information.

2 cannot insert NULL or update to NUL 50 This error can happen in context of column OBJECT_TYPE of table HOST_OBJE
8 L 31 CT_LOCKS_BASE when a primary key including column OBJECT_TYPE is in pla
7 ce. The primary key implicitly results in NOT NULL constraints on all involved col
umns while at the same time column OBJECT_TYPE of the source monitoring vie
w M_OBJECT_LOCKS sometimes contains NULL values.
With all current SAP HANA Revision levels the statistics server table HOST_OBJ
ECT_LOCKS_BASE is delivered without primary key. You can manually drop it w
ith the following command:

ALTER TABLE _SYS_STATISTICS.HOST_OBJECT_LOCKS_BASE DROP PRIMARY KEY

3 unique constraint violated 50 See "What can I do if the history collection fails with "301: unique constraint viola
0 15 ted"?" above.
1 50
16
50
17
50
34
an
do
the
rs

3 division by zero undefined: search tabl This error can be a consequence of check thresholds being erroneously set to 0. Se
0 e error: [6859] AttributeEngine: divid e SAP Note 2548692 for details. Be aware that also alert IDs different from 3 can
4 e by zero be responsible for the issue.

3 invalid number: not a valid number str 98 This error can happen for action 98 (ALERT_CHECK_CS_LOB_SPACE_RECLAI
3 ing MS) in case a parameter like garbage_collect_daily_schedule_s or garbage_collec
9 t_interval_s is set to a non-number value, e.g. the empty string. Make sure that th
ese parameters are properly configured, i.e. unset or set in line with the best pract
ices in SAP Note 2600030.

Issue number 298734 exists to make sure that the alert implementation gets smar
ter and uses M_CONFIGURATION_PARAMETER_VALUES instead of M_INIFI
LE_CONTENTS for the parameter check. M_CONFIGURATION_PARAMETER_
VALUES properly maps the empty string to the default value of 43200.
4 internal error: Failed to execute create 50 This error can happen with SAP HANA <= 2.00.059.04 and <= 2.00.063 for colle
0 and insert statement: General error: 1 49 ctors COLLECTOR_HOST_LOAD_HISTORY_HOST and COLLECTOR_HOST_
3 29 transaction rolled back by an intern 50 LOAD_HISTORY_SERVICE when temporary tables need to be created on second
al error: cannot change this transactio 50 ary system replication sites in context of join relocation. See SAP Note 3221290 fo
n's access mode from read-only to upd r more details and possible workarounds and solutions.
ate directly: please use "SET TRANSA
CTION READ WRITE" statement first

5 not supported type conversion This error can be a consequence of check thresholds being erroneously set to a no
9 n-integer value. The actual problem value is part of the reported error, e.g.:
2
Not supported conversion from NString to BigInt
('3749.000000000000000000' isn't a valid numeric value)

In this case a threshold for one alert was erroneously set to 3749.000000000000
000000. A change to 3749 resolved the issue.

You can find the configured thresholds in table _SYS_STATISTICS.STATISTICS_

ALERT_THRESHOLDS.

12 no data found This error can happen when alerts are configured inconsistently. Check metadata
9 tables like _SYS_STATISTICS.STATISTICS_ALERT_THRESHOLDS and correct
9 existing mistakes (e.g. missing severity levels).

6 missing smtp configuration in STATIS See SAP Note 2529478 and make sure that the SMTP parameters are set based on
0 TICS_PROPERTIES your needs:
0
0 internal.smtp.port
internal.smtp.sender
internal.smtp.server
 <action> is disabled and will not be ca var This message is recorded when SAP HANA has disabled a check due to a long run
lled iou time, see "Why are some statistics server actions disabled?" above for more detail
s s.

If you want to understand why the long runtime happened, you can analyze the pe
rformance and activities of SAP HANA during the relevant time frame:

Start time: Timestamp mentioned in alert minus check interval of impacted

action
End time: Timestamp mentioned in alert

Example:

Detail Value

Alert 603 is disabled and will not be called.

Alert timestamp 2016/02/24 09:25:14

Check interval of action 603 60 seconds

Critical time frame 2016/02/24 09:24:14 - 2016/02/24 09:25:14

Attention: The timestamp shown in the error message (e.g. "12 is disabled and w
ill not be called. Reason: timeout SNAPSHOT_ID: 2017-09-13 11:23:50.0000000
.") is UTC time, not local system time.

Attention: STATISTICS_ALERTS.ALERT_TIMESTAMP contains the UTC time

stamp rather than the local system timestamp in older SAP HANA Revisions, so y
ou should better check STATISTICS_ALERTS_BASE.

Typical reasons for timeouts are:

General system performance issues (SAP Note 2000000)
Crash of SAP HANA services (SAP Note 2177064)
Solution Manager requests blocking standard statistics server runs with rec
ord locks (SAP Note 1999998) on STATISTICS_SCHEDULE table: can be el
iminated with new Solution Manager monitoring for SAP HANA (SAP Note
2374272) available for SAP HANA 1.00.122.04 and higher
Secondary symptom of "feature not supported: invalid character encoding"
errors related to check ID 5034 (see above)
Secondary symptom of "Not supported conversion from NString to Integer"
for large threshold values for action ID 27 (SAP Note 2962294)
Secondary symptom of "Invalid forwarding to datacenter" that can happen f
or action IDs like 21, 30, 54, 78 and 89 that implicitly access the secondary s
ystem replication site. See SAP Note 2380176 for more details about this err
or scenario.
Timeouts for ID 5032 can be caused by a high amount of "(unknown)" objec
ts in HOST_OBJECT_LOCK_STATISTICS_BASE. These entries can be dele
ted as described in question "How can the memory requirements of the stati
stics server be minimized?" of this FAQ.

21. What happens if the STATISTICS_SCHEDULE table is manually modified in a wrong way?
Sometimes updates on the STATISTICS_SCHEDULE table are necessary to adjust settings. Erroneous updates can have
undesired effects, so you have to make sure that you only adjust the table in a correct way. For example, setting the STATE to
a value different from 'Disabled', 'Idle', 'Inactive' or 'Scheduled' means that the action will no longer be executed. You can use
SQL: "HANA_StatisticsServer_Schedule" (ONLY_UNKNOWN_STATES = 'X') available via SAP Note 1969700 to detect
faulty states.

22. How can I analyze and resolve SAP HANA alerts?

See SAP Note 2445867 for instructions how to analyze and resolve SAP HANA alerts.
23. Is it possible to restart the statistics server?
In general the statistics server is restarted when the related SAP HANA service (indexserver or nameserver in MDC system
DB) is restarted. For some reasons it is helpful to restart the statistics server individually, e.g.:
Statistics server worker threads disappeared and so the statistics server no longer works
Statistics server runs in statement memory limit OOMs due to a high amount of erroneously booked context memory
(SAP Note 2584388)
This restart is possible by temporarily setting the statistics server thread pool to 0:

<service>.ini -> [statisticsserver] -> threadpool = 0

See SAP Note 2584388 for more details.

See also SAP Note 2658611 that describes additional approaches to restart and re-initialize statistics server components.

24. What is the difference of SNAPSHOT_ID and SERVER_TIMESTAMP in histories?

Statistics server histories typically contain the columns SNAPSHOT_ID and SERVER_TIMESTAMP, both containing a
timestamp. The reasoning behind and the main differences are:
SNAPSHOT_ID
Scheduled start time of the history collection
Mainly used for statistics server internal purposes
May be used as part of a unique index
UTC time to avoid duplicates in context of summer / winter time changes
Identical for all histories being collected in one run
SERVER_TIMESTAMP
Time when history data was actually collected (typically shortly after the scheduled start time of the history
collection)
Used for tool-based history evaluation (e.g. SQL statements available via SAP Note 1969700)
Local system time
Can vary between different histories collected in one run

25. Why do different statistics server actions run in the system database and tenants?
In multitenant environments (SAP Note 2101244) different statistics server actions are run in the system database and the
tenants based on predefined best practices. It is controlled by the internal.sizing.profile entry in the statistics server control
table STATISTICS_PROPERTIES (SAP Note 2529478). Adjusting this parameter is possible, but it should normally not be
required.
In MCD system databases several alerts are deactivated per default for performance reasons so you shouldn't use the system
database for development or production operations. If you want to activate all alerts for specific reasons, you can adjust the
profile from SYSTEMDB to M:

UPDATE "_SYS_STATISTICS".STATISTICS_PROPERTIES SET VALUE = 'M' WHERE KEY = 'internal.sizing.profil

e'

When you want to return to the original state, you can switch back to SYSTEMDB:

UPDATE "_SYS_STATISTICS".STATISTICS_PROPERTIES SET VALUE = 'SYSTEMDB' WHERE KEY = 'internal.sizin

g.profile'

26. How can I monitor and tackle issues with event acknowledgement?
Events are an internal mechanism to inform SAP HANA components like the statistics server about issues, e.g. system
replication issues.
You can evaluate existing events via SQL: "HANA_Events" (SAP Note 1969700).
Check ID M0720 of the SAP HANA Mini Checks (SAP Note 1999993) reports an issue in case events are not purged in time.
In general events should be handled properly, exceptions should always be analyzed. The following known issues and
exceptions exist:
SAP HANA <= 1.00.102.00: General problems with event cleanup
SAP HANA >= 2.00.040: It is normal that a few AddressspaceUsage events exist on a permanent basis, so they can be
ignored.
It is possible to use SAP HANACleaner (SAP Note 2399996) in order to schedule an automatic cleanup of events.
Alternatively you can manually acknowledge the events using hdbcons (SAP Note 2222218). Use "-e <process>" or "-p <pid>"
to individually connect to the services with a high number of events in M_EVENTS:

hdbcons "event acknowledge -t AddressspaceUsage"

hdbcons "event acknowledge -t DiskFull"
hdbcons "event acknowledge -t ReplicationError"
hdbcons "event acknowledge -t SRConfigurationParameterMismatch"
hdbcons "event acknowledge -t SRConnectionClosed"

Afterwards you can delete the acknowledged events:

ALTER SYSTEM DELETE ALL HANDLED EVENTS

27. How can I check the state of the statistics server and restart it if required?
SAP Note 2896127 describes how to check the state of the statistics server and steps to activate it again.

28. Can Native Storage Extension (NSE) be used in context of the statistics server tables?
When Native Storage Extension (NSE, SAP Note 2799997) is used, it is recommended to configure it also for the column store
tables of the statistics server. See SAP Note 2983008 for more details. The following parameter is adjusted when NSE is
activated (true) or deactivated (default: false)

indexserver.ini -> [statisticsserver] -> page_loadable

29. Are there SQLScript version requirements for the statistics server?
For SAP HANA >= 2.0 the statistics server uses SQLScript version 3 (v3). If for some reasons (e.g. SAP Note 2924907) an
earlier SQLScript version is configured (indexserver.ini -> [sqlscript] -> kernel_mode), the statistics server may terminate
with unspecific errors like "Error during installation: SQL error $code$: $text$".

30. Do statistics server histories also contain information for remote system replication sites?
Up to SAP HANA 2.0 SPS 05 the statistics server histories only contain information for the local database. Starting with SAP
HANA 2.0 SPS 06 and in SAP HANA Cloud also data of other system replication sites (SAP Note ) is collected for most
histories (and some alerts). The system replication site is indicated in column SITE_ID of the history tables. For technical
reasons the following histories don't contain information of other system replication sites:
Consistency checks (GLOBAL_CATALOG_CONSISTENCY, GLOBAL_TABLE_CONSISTENCY)
SQL cache (HOST_SQL_PLAN_CACHE, HOST_SQL_PLAN_CACHE_OVERVIEW)
Job history (HOST_JOB_HISTORY)
Analysis commands of SAP Note 1969700 are provided in specific versions for SAP HANA >= 2.0 SPS 06 that include the
possibility to filter and / or aggregate the system replication site.
You can use SQL: "HANA_StatisticsServer_Schedule" (SAP Note 1969700) in order to determine to what extent statistics
server actions consider remote system replication sites. Depending on the value in column EXT_TYPE the behavior is as
follows:
EXT_TYPE = '': Only primary site is considered
EXT_TYPE = 'SR': All kinds of remote system replication sites are considered
EXT_TYPE = 'AA': Remote system replication sites of type Active/Active (read enabled) are considered

Keywords
GLOBAL_DEC_EXTRACTOR_STATUS
GLOBAL_DISKS
GLOBAL_INTERNAL_EVENTS
GLOBAL_PERSISTENCE_STATISTICS
GLOBAL_ROWSTORE_TABLES_SIZE
GLOBAL_TABLE_PERSISTENCE_STATISTICS
HOST_BLOCKED_TRANSACTIONS
HOST_COLUMN_TABLES_PART_SIZE
HOST_CONNECTIONS
 HOST_CONNECTION_STATISTICS
HOST_CS_UNLOADS
HOST_DATA_VOLUME_PAGE_STATISTICS
HOST_DATA_VOLUME_SUPERBLOCK_STATISTICS
HOST_DELTA_MERGE_STATISTICS
HOST_HEAP_ALLOCATORS
HOST_LONG_IDLE_CURSOR
HOST_LONG_RUNNING_STATEMENTS
HOST_LONG_SERIALIZABLE_TRANSACTION
HOST_OBJECT_LOCKS
HOST_OBJECT_LOCK_STATISTICS
HOST_ONE_DAY_FILE_COUNT
HOST_RECORD_LOCKS
HOST_RESOURCE_UTILIZATION_STATISTICS
HOST_RS_MEMORY
HOST_SAVEPOINTS
HOST_SERVICE_COMPONENT_MEMORY
HOST_SERVICE_MEMORY
HOST_SERVICE_STATISTICS
HOST_SERVICE_THREAD_SAMPLES
HOST_SQL_PLAN_CACHE
HOST_SQL_PLAN_CACHE_OVERVIEW
HOST_UNCOMMITTED_WRITE_TRANSACTION
HOST_VOLUME_FILES
HOST_VOLUME_IO_DETAILED_STATISTICS
HOST_VOLUME_IO_RETRY_STATISTICS
HOST_VOLUME_IO_TOTAL_STATISTICS
HOST_WORKLOAD
STATISTICS_ALERTS
STATISTICS_ALERT_INFORMATION
STATISTICS_ALERT_THRESHOLDS
STATISTICS_EMAILRECIPIENTS
STATISTICS_EMAIL_PROCESSING
STATISTICS_OBJECTS
STATISTICS_PROPERTIES
STATISTICS_SCHEDULE
is disabled and will not be called

Attributes
Key Value

Other Components SAP HANA > SAP HANA Database (HAN-DB)

Products
Products

SAP HANA, platform edition all versions

This document refers to

SAP Component Title
Note/KBA

2970921 HAN-DB-MON

2962294 HAN-DB- HANA Alert 27 is disabled and will not be called. Reason: timeout SNAPSHOT_ID: <timestamp>
MON
2800006 HAN-CPT- FAQ: SAP HANA Cockpit
CPT2

2799997 HAN-DB-NSE FAQ: SAP HANA Native Storage Extension (NSE)

2600030 HAN-DB Parameter Recommendations in SAP HANA Environments

2548692 HAN-DB- Alert in HANA Studio '[3] 304 division by zero undefined: search table error: [6859] AttributeEngine:
MON divide by zero'

2529478 HAN-DB How-To: Configuring SAP HANA Statistics Server Parameters

2505218 HAN-DB-BAC Large Log Backups due to large backup catalog

2445867 HAN-DB- How-To: Interpreting and Resolving SAP HANA Alerts

MON

2399996 HAN-DB- How-To: Configuring automatic SAP HANA Cleanup with SAP HANACleaner
MON

2380176 HAN-DB FAQ: SAP HANA Database Trace

2347981 HAN-DB- TrexUpdate failed on table '_SYS_STATISTICS:HOST_JOB_HISTORY_BASE'

MON

2239843 HAN-DB Error: Collector_Host_Sql_Plan_Cache: invalid column name: AVG_CHILD_THREAD_COUNT

2223237 HAN-DB Troubleshooting HANA Embedded Statistics Server Migration - decision tree

2222250 HAN-DB- FAQ: SAP HANA Workload Management

PERF

2222218 HAN-DB FAQ: SAP HANA Database Server Management Console (hdbcons)

2222110 HAN-DB- FAQ: SAP HANA Load History

PERF

2186744 HAN-DB FAQ: SAP HANA Parameters

2185556 HAN-CPT- FAQ: SAP HANA Cockpit (delivered with SAP HANA 1.0)
ADM

2180119 HAN-DB-SDA FAQ: SAP HANA Smart Data Access

2177064 HAN-DB- FAQ: SAP HANA Service Restarts and Crashes

ENG

2159435 HAN-DB How-To: Keeping SAP HANA Row Store in Memory when restarting

2143679 HAN-DB- How-To: Removing Primary Keys of SAP HANA Statistics Server Histories
MON

2137142 HAN-DB- HANA Error: 301 unique constraint violated: TrexUpdate failed on table
MON "_SYS_STATISTICS:HOST_SERVICE_THREAD_SAMPLES_BASE"

2133799 HAN-DB- Analyzing problems with delivery of e-mails for Alerts raised by the SAP HANA Embedded
MON Statisticsserver checks

2127852 HAN-DB- HANA Error: 301 unique constraint violated: TrexUpdate failed on table
MON

2116157 HAN-DB FAQ: SAP HANA Consistency Checks and Corruptions

2114710 HAN-DB- FAQ: SAP HANA Threads and Thread Samples

MON

2113228 HAN-DB- SAP HANA: Embedded statistics server is not correctly purging
MON

2101244 HAN-DB FAQ: SAP HANA Multitenant Database Containers (MDC)

2084747 HAN-DB How-To: Disabling Memory-intensive Data Collections of standalone SAP HANA Statistics Server

2081609 HAN-DB- How to Handle HANA Alert 0: 'Internal statistics server problem'
MON
2044468 HAN-DB- FAQ: SAP HANA Partitioning
ENG

2036630 HAN-DB- statisticsserver migration fails with 'error installing' / 'SQL error 129'
MON

2000002 HAN-DB- FAQ: SAP HANA SQL Optimization

PERF

2000000 HAN-DB- FAQ: SAP HANA Performance Optimization

PERF

1999998 HAN-DB- FAQ: SAP HANA Lock Analysis

PERF

1999993 HAN-DB- How-To: Interpreting SAP HANA Mini Check Results

MON

1999930 HAN-DB FAQ: SAP HANA I/O Analysis

3221290 HAN-DB Internal Statistical Transaction is Rolled Back Due to Failure to Create a Temporary Table for Join
Relocation

2983008 HAN-DB Enable/Disable NSE (PAGE LOADABLE) for column tables of HANA statistics server

2924907 HAN-DB Unexpected Results in SQLScript of SAP HANA 2 SPS04 when Variables with Same Prefix are Used

2896127 HAN-DB HANA Statistics Server – Checking the status

2658611 HAN-DB- Embedded Statistics Service Control in live operation

MON

2584388 HAN-DB High Memory Usage in Allocator Connection/XXXXXX/Statement/YYYYYYYY/IMPLICIT by User

_SYS_STATISTICS

2478767 HAN-DB Selecting Service Thread Sample Data Fails With an "invalid character encoding" Error Due to Columns
PASSPORT_ROOTCONTEXT_ID or PASSPORT_CONNECTION_ID

2453766 HAN-DB Selecting Service Thread Sample Data Fails With an "invalid character encoding" Error Due to Columns
PASSPORT_ACTION or PASSPORT_COMPONENT_NAME

2374272 HAN-DB- Enabling new HANA Monitoring mechanism for Solution Manager/Focused Run
MON

2092033 HAN-DB- Embedded Statistics Service Migration Guide

MON

1969700 HAN-DB SQL Statement Collection for SAP HANA

1917938 HAN-DB- Migration of the statistics server for Revision 74 or higher

ENG

Embedded Statistics Service Workload Class

Embedded Statistics Server Views

SAP HANA Administration Guide

This document is referenced by

SAP Component Title
Note/KBA

3583835 HANA_Tables_TopGrowingTables_Size_History/HANA_Tables_TopGrowingTables_Records_History
only shows the old data.

3307159 HAN-DB- DB statistics server is not working

MON

3217863 HAN-DB- Alert "Statistics server has been disabled. Data might be out of date"
MON
2800030 HAN-DB- FAQ: SAP HANA Kernel Profiler
PERF

1999998 HAN-DB- FAQ: SAP HANA Lock Analysis

PERF

2697558 HAN-DB- Increased Memory Consumption in Allocator M_DEV_BACKUP_CATALOG_LOG_FILES

BAC

2800055 HAN-DB FAQ: SAP HANA Kernel Sentinel

3096705 HAN-DB- Regular high CPU usage caused by statistics server

PERF

2600076 HAN-DB FAQ: SAP HANA Inverted Individual Indexes

2600030 HAN-DB Parameter Recommendations in SAP HANA Environments

3030821 HAN-DB- Cockpit does not show longer history than 10 days for performance monitor history data.
MON

2800020 HAN-DB FAQ: SAP HANA Triggers

2766055 HAN-DB- SAP HANA Statistics Server issue

MON

2943943 HAN-DB- Outdated DiskFull with state HANDLED Alerts

MON

2943850 HAN-DB- HANA alerts show incorrect date and / or time

MON

2160391 HAN-DB- FAQ: SAP HANA Indexes

ENG

2642446 BC-DB-HDB- Transaction DBACOCKPIT gives a warning that the "statisticsserver is deactivated"
CCM

2908132 HAN-DB- How to change the threshold value or scheduling interval for HANA Alerts
MON

2215093 HAN-DB- After upgrading the HANA Database the statisticsserver does not work
MON

2889270 HAN-DB- High CPU Usage by Embedded Statistics server with CALL
MON _SYS_STATISTICS.STATISTICS_SCHEDULABLEWRAPPER

2852678 BC-DB-HDB- Check for disabled Alerts in Self-Monitoring

CCM

2800028 HAN-DB FAQ: SAP HANA Optimizer Statistics

2772514 HAN-DB- EWA report for HANA has the alert 'Identifies internal statistics server problem'
MON

2154844 HAN-DB HANA Alerts keep getting disabled after switching to the embedded statisticserver

2169283 HAN-DB- FAQ: SAP HANA Garbage Collection

ENG

2638539 HAN-DB- Difference between Statisticsserver check status 'Inactive' and 'Disabled'
MON

2736234 HAN-LM- Upgrade from HANA 1.0 to 2.0 failed due to ESS not converted
UPG-DB

2732928 HAN-DB-HA How to deal with Alert "Service on <hostname>:<service port number> has increased log replay backlog"

2313619 HAN-DB- How-To: Generating and Evaluating SAP HANA Call Stacks
MON

2691613 HAN-DB- Uncommon growth of "_SYS_STATISTICS"."HOST_VOLUME_FILES_BASE" table

MON
2656792 HAN-DB- Alert [6000] missing smtp configuration in STATISTICS_PROPERTIES SNAPSHOT_ID...
MON

2629590 HAN-DB- Recommendation for Interval Length for a statistics server collector
MON

2593571 BC-DB-LCA FAQ: SAP HANA Integrated liveCache

2536974 HAN-DB- The system is running without alerts, but not all checks are being performed
MON

2535951 HAN-DB- FAQ: SAP HANA Users and Schemas

SEC

2186744 HAN-DB FAQ: SAP HANA Parameters

2529478 HAN-DB How-To: Configuring SAP HANA Statistics Server Parameters

2493696 HAN-DB- EWA Report - "Number of collectors with retention times < 14" is equal to 1
MON

2488311 HAN-DB- How to Manage the Volume of Data Collected by the Statistics Service
MON

2124112 HAN-DB- FAQ: SAP HANA Parsing

PERF

2477204 HAN-DB FAQ: SAP HANA Services and Ports

2399996 HAN-DB- How-To: Configuring automatic SAP HANA Cleanup with SAP HANACleaner
MON

2400024 HAN-DB- How-To: SAP HANA Administration and Monitoring

MON

2222277 HAN-DB- FAQ: SAP HANA Column Store and Row Store
PER

2388483 HAN-DB How-To: Data Management for Technical Tables

1999880 HAN-DB-HA FAQ: SAP HANA System Replication

2380176 HAN-DB FAQ: SAP HANA Database Trace

2222110 HAN-DB- FAQ: SAP HANA Load History

PERF

2222218 HAN-DB FAQ: SAP HANA Database Server Management Console (hdbcons)

2222250 HAN-DB- FAQ: SAP HANA Workload Management

PERF

2222220 HAN-DB- FAQ: SAP HANA DBACOCKPIT

ENG

2073112 HAN-STD- FAQ: SAP HANA Studio

ADM-DBA

2116157 HAN-DB FAQ: SAP HANA Consistency Checks and Corruptions

2143679 HAN-DB- How-To: Removing Primary Keys of SAP HANA Statistics Server Histories
MON

2000000 HAN-DB- FAQ: SAP HANA Performance Optimization

PERF

2114710 HAN-DB- FAQ: SAP HANA Threads and Thread Samples

MON

2159435 HAN-DB How-To: Keeping SAP HANA Row Store in Memory when restarting

1999020 HAN-DB How-To: SAP HANA Troubleshooting when Database is no longer accessible

2000003 HAN-DB- FAQ: SAP HANA

ENG
1999993 HAN-DB- How-To: Interpreting SAP HANA Mini Check Results
MON

2000002 HAN-DB- FAQ: SAP HANA SQL Optimization

PERF

1803039 HAN-DB Alert 0 Internal error for Monitor CHECK_HOSTS_CPU

2997712 HAN-DB SAP HANA 2 SPS04 Database Maintenance Revision 048.04

3004477 HAN-DB Usage of statistics server test alert (ID 999)

2983008 HAN-DB Enable/Disable NSE (PAGE LOADABLE) for column tables of HANA statistics server

2957968 SV-SMG- Activate Table Statistics on HDB for DVM Cloud Application
DVM

2915604 HAN-DB Indexserver Crashes on ptime::ExternalStatement::ExecutionResponse::serialize on Worker Node of

Scale Out System

2846349 HAN-DB-HA Service Crash After federation::FederationContext::rdes_for_props

2570661 HAN-DB Exception 71000287 During Execution of Internal Statement Within Procedure
"_SYS_STATISTICS"."HELPER_ALERT_MISSING_VOLUME_FILE_AGE"

2443239 HAN-DB Statistics Service is Disabled After Upgrade or Takeover

2461913 HAN-DB Indexserver Crashes During Startup With "Assertion failed: size_t(readStart) - 1 ==
m_RowMapToTarget.size()"

1977584 HAN-DB Technical Consistency Checks for SAP HANA Databases

2260972 HAN-DB- High memory consumption if some checks of embedded Statistics Server are called manually
MON

1969700 HAN-DB SQL Statement Collection for SAP HANA

706478 HAN-DB Preventing Basis tables from increasing considerably