SAP HANA Lock Analysis: Problems and Solutions

by ·

1999998 – FAQ: SAP HANA Lock Analysis




Symptom

SAP HANA connections are hanging because they have to wait for locks.

Environment

SAP HANA

Cause

1. Where do I find information about SAP HANA lock analysis?
2. Which indications exist for SAP HANA locking issues?
3. What types of locks exist?
4. How can I check if my SAP HANA database suffers from lock waits?
5. How can transactional lock waits be analyzed and optimized?
6. How can internal lock waits be analyzed?
7. How can special lock waits be analyzed?
8. Do timeouts for lock waits exist?
9. What happens in case of deadlocks?
10. Is it possible to identify the record responsible for a record lock?
11. How can I determine how long certain types of locks are held?
12. What is the unit of column TOTAL_LOCK_WAIT_DURATION in monitoring view M_SQL_PLAN_CACHE?
13. How can I automatically capture additional information in case of unusual locking issues?
14. Why are there deadlocks although changes are done in a sorted manner?
15. How can the lock holder be determined?

Resolution

1. Where do I find information about SAP HANA lock analysis?

SAP Note 1858357 provides an overview how to analyze lock situations.

The SAP HANA Troubleshooting and Performance Analysis Guide at SAP HANA Troubleshooting and Performance Analysis Guide contains further information related to lock analysis.

2. Which indications exist for SAP HANA locking issues?

You suffer from terminations and short dumps due to SQL errors 131 or 133:

SQL error 131: transaction rolled back by lock wait timeout
SQL error 133: transaction rolled back by detected deadlock
SQL error 146: resource busy and NOWAIT specified

The trace files contain messages like:

There are too many lock items on this system.

Deadlock detected: Deadlock detected while executing transaction (TRANSACTION_ID=<tid>, UPDATE_TRANSACTION_ID=<utid>): 
This is not an HDB error. User or application may cause a deadlock due to incorrect access sequences on shared objects.

The following SAP HANA alerts indicate problems in the locking area:

Alert Name Description
16 Lock wait timeout configuration Determines whether the ‘lock_wait_timeout’ parameter in the ‘transaction’ section of the indexserver.ini file is between 100,000 and 7,200,000.
49 Long-running blocking situations Identifies long-running blocking situations.
59 Percentage of transactions blocked Determines the percentage of transactions that are blocked.

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
870 SelfWatchDog activity time (%, last hour)
1010 Age of oldest active trans. lock wait (s)
1011 Trans. lock wait durations > 600 s (last day)
1020 Threads currently waiting for locks
1021 Maximum threads waiting for locks (last day)
1030 Concurrently blocked transactions (last hour)
1031 Concurrently blocked transactions (last day)
1040 Total current record locks
1045 Transactional locks older than 1 day
1050 Significant internal lock waits (last hour)
1052 Significant internal lock waits (last day)
1720 Supported nameserver lock file location

3. What types of locks exist?

We can distinguish the following SAP HANA lock types:

Lock type Lock wait thread state Scope  Views Details
Record lock ConditionalVariable Wait
(RecordLockWaitCondStat / TransactionLockWaitCondStat)		 transactional M_RECORD_LOCKS
M_CONDITIONAL_VARIABLES
M_BLOCKED_TRANSACTIONS
M_OBJECT_LOCK_STATISTICS		 Exclusive locks on record level, typically caused by concurrent changes of the same records by different transactions
Object lock ConditionalVariable Wait
(TableLockWaitCondStat / TransactionLockWaitCondStat)		 transactional M_OBJECT_LOCKS
M_CONDITIONAL_VARIABLES
M_BLOCKED_TRANSACTIONS
M_OBJECT_LOCK_STATISTICS		 Locks on object level, typically caused by DDL operations requiring an object lock:

  • Shared locks (INTENTIONAL EXCLUSIVE): Set in case of DELETE, INSERT, MERGE, SELECT FOR UPDATE, UPDATE and UPSERT operations on table
  • Exclusive locks (EXCLUSIVE): Set in case of DDL operations on table and in case of an explicit LOCK TABLE operation
Metadata lock ConditionalVariable Wait
(TransactionLockWaitCondStat)		 transactional M_CONDITIONAL_VARIABLES
M_BLOCKED_TRANSACTIONS		 Metadata locks
Read / write lock ExclusiveLock Enter
IntentLock Enter
SharedLock Enter		 internal M_READWRITELOCKS Read / write lock waits, e.g. waits during critical savepoint phase
Barrier Semaphore BarrierSemaphore Wait internal
Semaphore Semaphore Wait internal M_SEMAPHORES Low level locks based on semaphores, e.g. waits for a critical delta merge phase
Mutex Mutex Wait internal M_MUTEXES Low level locks based on mutexes
Barrier Barrier Wait internal Low level locks similar to mutexes
Speculative locks Speculative Lock Retry backoff
Speculative Lock Wait for fallback		 internal Low level transactional memory locks (SAP HANA >= Rev. 122)
liveCache lock internal M_LIVECACHE_LOCKS Locks related to integrated liveCache (if used)
Nameserver topology special The nameserver topology is locked based on a file (/tmp/.hdb_<sid>_<inst_id>_lock)

In addition to these explicit locks there can be some “hidden” lock waits of threads in status ‘Running’. Related call stacks (SAP Note 2313619) contain:

  • __GI___sched_yield
  • Synchronization::impl::SpinLock::lock
  • ptime::Futex::lock
  • Synchronization::SystemMutex::lock
  • __lll_lock_wait / pthread_mutex_lock (Linux kernel lock)

Record and object locks are mainly linked to the application transactions while read / write locks, semaphores, mutexes and barriers are managed by SAP HANA internally.

Be aware that the table above concludes from a lock type to a thread state. The opposite way is not generally valid. For example, there can be “ConditionalVariable Wait” situation that are not linked to a record lock or object lock.

4. How can I check if my SAP HANA database suffers from lock waits?

On a very elementary level you can identify lock waits based on the thread states.

Current threads can be displayed via:

  • Transaction DBACOCKPIT: Performance –> Threads
  • SAP HANA Studio: Performance –> Threads
  • SQL: “HANA_Threads_CurrentThreads” (SAP Note 1969700)

Historic thread activities can be determined via:

  • SQL: “HANA_Threads_ThreadSamples_FilterAndAggregation” and “HANA_Threads_ThreadSamples_AggregationPerTimeSlice” (SAP Note 1969700)

The following thread states indicate lock wait situations:

  • Barrier Wait
  • ConditionalVariable Wait
  • ExclusiveLock Enter
  • Mutex Wait
  • Semaphore Wait
  • SharedLock Enter
  • Sleeping

Not every occurrence of these states is critical. It happens frequently that a thread submits a request to another thread and waits for a related semaphore until the called thread has returned the result.

The comprehensive check command SQL: “HANA_Configuration_MiniChecks” (SAP Note 1969700) contains also lock related checks. An ‘X’ in column C indicates potentially critical situations. Example output:

lockChecks.JPG

The following SQL errors also indicate lock related problems:

  • SQL error 131: transaction rolled back by lock wait timeout
  • SQL error 133: transaction rolled back by detected deadlock

The following SAP HANA alerts are another indication for locking issues:

  • Alert 49: Long-running blocking situations
  • Alert 59: Percentage of transactions blocked

Below you can find more specific information how to analyze certain lock wait situations.

5. How can transactional lock waits be analyzed and optimized?

If a high number of transactional (record and object) lock waits exist, but at the same time there are also many semaphore and mutex waits, we can usually assume that the record and object lock waits are a consequence of the mutex and semaphore waits. In this case you should focus on the analysis of mutex and semaphore waits (see below). Otherwise you can use the following approaches to analyze transactional lock situations.

Current lock wait situations are visible as blocked transactions:

  • View M_BLOCKED_TRANSACTIONS
  • SQL: “HANA_Locks_Transactional_LockWaits[_Hierarchy]” (DATA_SOURCE = ‘CURRENT’) available via SAP Note 1969700
  • SAP HANA Studio: Performance -> Blocked Transactions
  • DBACOCKPIT: Diagnostics -> Locks -> Blocked Transactions

Historic lock wait situations can be found in the related history:

  • View HOST_BLOCKED_TRANSACTIONS
  • SQL: “HANA_Locks_Transactional_LockWaits”[_Hierarchy] (DATA_SOURCE = ‘HISTORY’) and SQL: “HANA_Locks_Transactional_LockWaits_PerObject” available via SAP Note 1969700
  • SQL: “HANA_LoadHistory_Services” (SAP Note 1969700)

Among others this information provides information about the involved table and transactions. With some further research it is usually possible to identify the related business scenarios.

The following table lists typical approaches to optimize record and object locks:

Lock type Optimization approaches
Record lock Check from an application and / or scheduling perspective if you can reduce the amount of concurrent changes of the same records by different transactions.

Check from application side if you can reduce the critical time frame between DML operation and next COMMIT (lock is always held until next COMMIT is executed).

Manually terminate the lock holding transaction if it is responsible for a critical current problem.

Lock wait related recommendations for specific tables like REPOSRC, FKKDIHDTMP, NRIV or /PLMB/FRW_BUFFER can be found in SAP Note 2000002 (“Are there standard recommendations for specific SQL statements available?”).
Object lock Make sure that critical DDL operations (e.g. offline repartitioning) are executed at times with minimum concurrent workload.

Check if you can reduce the amount of critical DDL operations (e.g. by using online operations).

If another transaction holds an uncommitted change on a table, a DDL operation on the same table like CREATE INDEX will escalate this row level lock to a table level lock because it prevents any new DML operation to be started until it acquires the lock itself and finalizes its tasks. In order to minimize this risk, you can reduce the lock wait timeout on transaction level by executing the following command before starting the DDL operation:

SET TRANSACTION LOCK WAIT TIMEOUT <timeout_in_ms>

A value of 0 for <timeout_in_ms> would result in an immediate termination of the DDL operation with the following error:

SQL error 131: transaction rolled back by lock wait timeout: Lock timeout occurs while waiting OBJECT_LOCK of mode EXCLUSIVE

For index creations additional options exist to reduce the risk and duration of object locks. See SAP Note 2160391 (“Are accesses to the underlying table locked when an index is created?”) for more information. If an index is created with NOWAIT option and there is still an open change on the table, it is terminated with the following error:

SQL error 146: resource busy and NOWAIT specified

Manually terminate the lock holding transaction if it is responsible for a critical current problem.
Metadata lock This kind of locking situation in combination with thread method “Authentication” can happen if many sessions connect to the SAP HANA database at the same time. Typically this scenario is a victim of another, major problem. The risk of running into this scenario can be reduced by adjusting the SAP HANA parameter indexserver.ini -> [authentication] -> last_successful_connect_update_interval. See SAP Note 2460123 for more details.

6. How can internal lock waits be analyzed?

Read / write lock, semaphore and mutex waits are typically related to internal lock situations. If you experience a high number of threads waiting for these locks you can run SQL: “HANA_Sessions” (SAP Note 1969700) in the first step. The column WAITING_FOR can provide some additional expert insight for threads with states ‘SharedLock Enter’, ‘ExclusiveLock Enter’, ‘Mutex Wait’ and ‘Semaphore Wait’.

Example:

waiting_for.JPG

In this example the lock names already indicate a contention related to the nameserver process.

Similar information can be found in column LOCK_WAIT_NAME of views M_SERVICE_THREADS and M_SERVICE_THREAD_SAMPLES.

The following table contains typical general reasons for contention on internal locks:

Check Details
Memory bottleneck Check if there are indications for a memory bottleneck like paging or column store unloads at times of the internal lock waits. If yes, identify and eliminate the root cause of the memory bottleneck. See SAP Note 1999997 for more details.
CPU bottleneck Check if there are indications for a high CPU consumption at times of the internal lock waits. If yes, try to identify the root cause like expensive SQL statements (SAP Note 2000002) or CPU-intensive processes outside of SAP HANA. Be aware that sometimes the high CPU consumption is a consequence of the internal locks because of active waits.
SAP HANA bug Bugs like permanent deadlocks on internal resources can be responsible for internal lock waits. Therefore check SAP Notes for potential bugs.

Furthermore the following typical root causes exist for certain scenarios:

Thread state Lock name Reason
Barrier Wait Barrier waits don’t show a lock name, so it is sometimes hard to understand the root cause without further details, e.g. the call stack. The following typical scenarios exist:

Scenario Details
Thread method: ‘SessionControl’ These waits during an initial connection establishment are typically a consequence of another underlying issue, e.g. memory reclaim (SAP Note 1999997), waits for other internal locks or a resource bottleneck.
TransactionManager::TransactionControlBlockFactory::getTimestampState
TransactionManager::TransactionControlBlockSPI::waitStateChange:suspendForSynchronization		 These waits are linked to transaction management. They are typically a consequence of another underlying issue, e.g. memory reclaim (SAP Note 1999997), waits for other internal locks or a resource bottleneck.
sem_wait
Synchronization::SystemTimedSemaphore::wait
Execution::Thread::suspendForSynchronization
Synchronization::Barrier::wait
DataAccess::PersistenceSessionRegistry::registerSession
DataAccess::PersistenceSessionRegistry::unregisterSession		 These waits are linked to the iteration barrier related to the management of persistence sessions (see also PersistenceSessionRegistry mutex information below).

The iteration barrier is normally not allocated for a long time, but during the critical savepoint phase (SAP Note 2100009) the flush of private log buffers can allocate this barrier for a longer time. Therefore you have to make sure that this phase doesn’t take unnecessarily long, e.g. due to I/O problems or a very high number of existing threads.

 
ConditionalVariable Wait ParallelIndexHash_cond This lock wait is typically linked to thread method ‘HashDict worker’ and it impacts parallel job workers processing the same database request. See SAP Note 2114710 (Thread Method = ‘HashDict worker’) for more details.

Massive waits for this lock are typically a consequence of an expensive database request (SAP Note 2000002), involving intermediate results (dicts) in contexts like calculation scenarios or joins.

With SAP HANA Rev. <= 121 it can happen that after an out-of-memory situation threads permanently remain in a state waiting for ParallelIndexHash_cond. This bug is fixed with SAP HANA Rev. 122.
ExclusiveLock Enter
SharedLock Enter

ExclusiveLock Enter
SharedLock Enter

Semaphore Wait

Sleeping

 unnamed Mutex

BTree GuardContainer

SleepSemaphore

 The BTree GuardContainer lock synchronizes read and write accesses to individual columns in delta storage. Typically these locks happen on a single table at a time. Typical lock types and call stacks are:

Operations Thread state Lock name Call stack
DML ExclusiveLock Enter
Semaphore Wait		 unnamed Mutex
SleepSemaphore		 AttributeEngine::BtreeAttribute::GuardContainer::_acquire
AttributeEngine::BtreeAttribute::_acquireRetry (AttributeEngine::BtreeAttribute::GuardContainerLocks)
DML Sleeping AttributeEngine::BTreeAttribute<TrexTypes::StringAttributeValue>::commitOptimizeForStrings at DeltaLocks.h
AttributeEngine::BTreeAttribute<TrexTypes::StringAttributeValue>::getLastDocidsForStrings
SELECT SharedLock Enter BTree GuardContainer AttributeEngine::BtreeAttribute::GuardContainerLocks::acquireRead

Known situations are:

Patch level Context Details
Rev. 70 – 73 Join statistics creation When join statistics are created during the first execution of a join query, a deadlock with DML operations may happen on Rev. 70 to 73. See SAP Note 2009806 for more information.
SPS <= 09 Join with expensive sorting Up to SAP HANA SPS 09 sorting is secured by a GuardContainerLock, so you have to avoid joins with expensive sorting when it causes trouble (e.g. by using restricting selection conditions or by avoiding the ORDER BY). Alternatively you can upgrade to SAP HANA SPS 10 or higher.
Fulltext indexing Check if the underlying table has fulltext indexes (e.g. by using SQL: “HANA_Indexes_Columns”, SAP Note 1969700) and look at for threads in suspicious states like ‘indexing’ at the time of the locking issue (SAP Note 2114710). If fulltext indexes exist, you can suspend the indexing by executing the following command:

ALTER FULLTEXT INDEX "<schema_name>"."<index_name>" SUSPEND QUEUE

If the problem disappears, you can activate the queue during non-critical times in order to update the index using the following command:

ALTER FULLTEXT INDEX "<schema_name>"."<index_name>" ACTIVATE QUEUE

Be aware that the SUSPEND isn’t persisted, so after a SAP HANA restart the queue will be processed again and you have to executed the SUSPEND command another time to deactivate it.
Concurrent read and update transactions If a column is read and at the same time another transaction wants to change it, both the update transactions and all subsequent readers could be blocked until the reader is finished. Optimizing the critical read activities (SAP Note 2000002) can help to improve this situation.
Delta storage issues A particularly large delta storage or performance problems accessing information in the delta storage can also be responsible for these locks. You can check the delta merge configuration based on SAP Note 2057046.
Massively parallel changes If parallel change operations on the same table often wait in the following kinds of modules, you can consider an increase of package sizes / a reduction of COMMITs in order to reduce the amount of waits:

AttributeEngine::BTreeAttribute<TrexTypes::StringAttributeValue>::getLastDocidsForStrings
AttributeEngine::BTreeAttribute<TrexTypes::StringAttributeValue>::getLastDocids
AttributeEngine::AttributeApi::getLastDocids
TRexAPI::UdivLookup::map_dockeys_to_udivs
AttributeEngine::BTreeAttribute<TrexTypes::StringAttributeValue>::commitOptimizeForStrings
AttributeEngine::AttributeIndexJob::run
AttributeEngine::AttributeApi::commitOptimizeAttributes
Expensive queries Expensive queries can allocate the BTree GuardContainer lock in specific phases for a long time (particularly in combination with an increased delta storage), e. g.:

  • Accessing column information during a full table scan (call stack: *::mgetSearch, e.g. UnifiedTable::impl::VersionedNBitVectorDataObject::mgetSearch, UnifiedTable::NBitDataAccessor::mgetSearch, UnifiedTable::NBitDataAccessorUtils::mgetSearchRowPosSSE, sse_icc_lib::mgetSearchi_SSE4, sse_icc_lib::mgetSearchi_SSE4impl), see SAP Notes 2318281 and 2321573
  • Column document search (AttributeEngine::BTreeAttribute<TrexTypes::StringAttributeValue>::searchDocuments)

The best solution in this case is to tune these bad queries, so that both their own runtime and the effect on the overall system is minimized. See SAP Note 2000002 for more information related to SQL statement tuning.

In order to release a currently existing lock, you can terminate the related session as described in SAP Note 2092196.

 
Job Exec Waiting CSPlanExecutorWaitForResult This lock indicates that a SQL executor waits for its job workers executing a column store statement (thread method: PlanExecutor calc). So it is basically an “idle” lock and itself not critical. If the lock is visible for a long time, you have to check why the execution of the SQL statement by the job worker(s) takes so long.
Job Exec Waiting unnamed BinarySemaphore (SAP HANA 1.0)

JobBarrier Wait for Jobs (SAP HANA >= 2.00)

 These locks are visible when a thread has delegated the work to one or more parallel JobWorker threads and waits for the results. It can be considered as idle wait.
Mutex Wait AttributeStore Resource Load This lock is used when a database query requires a column which currently isn’t in memory and needs to be loaded into the column store. The actual load is done by special threads (LoadField) and the thread preparing or executing the query has to wait for the “AttributeStore Resource Load” lock during this time. See SAP Note 2127458 for more information about loads and unloads.
Mutex Wait ceManager_GCLock This lock indicates a wait for the calculation engine garbage collector (CalcEngineGarbageCollector, see SAP Note 2169283).
Mutex Wait ConfigurationRegistry This lock synchronizes concurrent changes of SAP HANA parameters (SAP Note 2186744). Related call stacks contain modules like:

  • Configuration::DefaultHandler::prepareReconfigure
  • config::ConfigMgr::reloadLayered
  • TrexService::CoreHandler::reconfigure
  • TrexService::HandlerRegistry::reconfigureAllHandler

The related thread detail is typically reconfigAll.

If the lock isn’t available after a few minutes, the change may be terminated with the following error:

129: transaction rolled back by an internal error
exception 70000030: error: general configuration error

If you see a high amount of these wait situations you should check if an unnecessary high amount of parameter changes is done or if a transaction got stuck while holding the lock.

A bug with SAP HANA Rev. <= 112.02 can be responsible for this locking issue.
Mutex Wait CreateDropSnapshotLock This lock is required when a snapshot activity (e.g. triggered by system replication or a backup) can’t be processed immediately. For example, a running savepoint may prevent snapshot activities. See SAP Note 2100009 for more information regarding savepoints and snapshots.
Mutex Wait CSPlanExecutorLock Job worker threads of a parallelized execution have to wait for other threads of the same connection. You can consider these waits as idle waits. In order to improve the performance and reduce the wait time, you have to check what the other threads of the same connection are doing and how their activity can be optimized (e.g. by eliminating lock scenarios, resource bottlenecks or expensive SQL statements).
Mutex Wait DeltaDataObject This lock is used to synchronize changes to the delta storage data page chain which includes adding new pages or rewrite of pages. The delta data page chain is used to persist rows inserted into delta. This lock is required during DML or DDL operations, but not when doing SELECTs.
Mutex Wait Federation This lock is linked to Smart Data Access (SDA), i.e. accesses to remote databases. See SAP Note 2180119 and check if you can optimize the configuration and use of SDA. Make sure that you use supported client software (e.g. ODBC driver).
Mutex Wait FixedDictFlush This lock is used to synchronize page dirty operations for fixed delta columns. This shows up during DML operations when new values are added to the fixed data type columns. It also shows up for every INSERT as the internal delta row id is also a fixed dictionary column.
Mutex Wait fo-Stats This lock is required to synchronize the collection of operations of a specific job graph (i.e. of a specific database request), so only job worker processes are impacted. It mainly protects adding new jobs.
Mutex Wait FunctionProfilerLock This lock is related to the function profiler trace which can e.g. be activated in the context of a performance trace (SAP Note 1787489).

With SAP HANA <= Rev. 102.01 this lock is also acquired when PlanViz (SAP Note 2073964) is used or when the related use_profiler parameter was deleted from the ini file rather than explicitly being set to false. These issues are fixed with SAP HANA Rev. 102.02.
Mutex Wait HugeAlignmentPool This lock can show up in case several threads execute specific memory allocations at the same time. A general memory bottleneck situation can increase the probability of these waits.
Mutex Wait Mutex[releasable_handle.hpp:202]: TransactionManager::WaitGraphImpl::WaitGraphImpl This lock is linked to the deadlock wait graph creation. With SAP HANA Rev. 110 to 112.02 it can be critical in the context of the SAP HANA bug described in SAP Note 2400515.
Mutex Wait JE_SynchronizedNTuple_Lock This lock synchronizes a parallelized creation of tuples during join operations. It only locks threads belonging to the same parallelized statement execution. If you repeatedly suffer from these lock waits, you may suffer from a bad SQL statement or an inefficiently processed join. See SAP Note 2000002 and check if you can optimize the underlying SQL statements.
Mutex Wait jx-pq00-nopref
jx-pq01-nopref
jx-pq02-nopref
jx-pq03-nopref
jx-pq04-nopref
jx-pq05-nopref
jx-pq06-nopref
jx-pq07-nopref

jx-pq00-nosteal
jx-pq01-nosteal
jx-pq02-nosteal
jx-pq03-nosteal
jx-pq04-nosteal
jx-pq05-nosteal
jx-pq06-nosteal
jx-pq07-nosteal

jx-pq00-numa
jx-pq01-numa
jx-pq02-numa
jx-pq03-numa
jx-pq04-numa
jx-pq05-numa
jx-pq06-numa
jx-pq07-numa

jx-pq00-system
jx-pq01-system
jx-pq02-system
jx-pq03-system
jx-pq04-system
jx-pq05-system
jx-pq06-system
jx-pq07-system

 These locks synchronize operations on job queues (separated by NUMA node and priority), so only job worker processes are impacted.

If you observe a high number of waits for this lock you can check if there are many expensive SQL statements (SAP Note 2000002) being processed with a high number of jobs.

The meanings of the suffixes are:

  • nopref: No restriction related to NUMA (non-uniform memory access) nodes
  • nosteal: Fix assignment to a specific NUMA node without ability to switch to another node
  • numa: Preference for a certain NUMA node, but possibility to run also on another node
  • system: Related to internal activities like garbage collection
Mutex Wait LastValuesHolder::AccessCoordinator This mutex synchronizes the accesses to the monitoring view M_STATISTICS_LASTVALUES, where the standalone statistics server stores data historicized during the most recent snapshot. The most efficient way to bypass this bottleneck is an upgrade to the embedded statistics server that is possible as of Rev. 74 (see SAP Note 1917938). Alternatively you can reduce the collected data volume as described in SAP Note 2084747.
Mutex Wait LoadLock This lock secures the load of delta storage information from disk. A significant amount of waits can indicate memory shortages (SAP Note 1999997) or disk I/O bottlenecks (SAP Note 1999930).
BarrierSemaphore Wait
Mutex Wait		 LoggerBufferSwitch This lock indicates wait situations for a log buffer switch. Typical reasons are:

Reason Details
Slow I/O writes to log area Check according to SAP Note 1999930 if there are general or temporary I/O write regressions to the local log area.
System replication performance See SAP Note 1999880 (“Can problems with system replication impact the performance on the primary system?”) and make sure that system replication is running smoothly without impacting the primary system.
Number and size of log buffers configured too small If none of the above reasons can explain the wait situations, you can consider to increase the number and the size of the log buffers. You can use SQL: “HANA_Logs_LogBuffers” (SAP Note 1969700) to check the current log buffer configuration. As a starting point you can double the configured size and number. The following parameters can be used to adjust size and number:

global.ini -> [persistence] -> log_buffer_size_kb
global.ini -> [persistence] -> log_buffer_count
Ad-hoc creation of new redo log files Redo log files that aren’t backed up or that are still required in context of savepoints or system replication can’t be re-used by SAP HANA. If no re-usable redo log exists, a new redo log file is created and initialized. This can take a certain amount of time – particularly in case of issues in the I/O area. During this time the redo log buffer can’t be flushed and LoggerBufferSwitch waits are possible.

The formatting of new redo log files is done by the LogSegmentFormatter[0] thread. The related call stack is:

__libc_pwrite64
System::UX::pwrite
FileAccess::fileWrite
FileAccess::LocalFile::write
DataRecovery::DirectoryLogSegment::formatStep
DataRecovery::DirectoryLogSegmentFactory::run

The following options exists to improve this situation:

  • Disk I/O tuning (SAP Note 1999930)
  • Elimination of bottlenecks in the area of savepoints (SAP Note 2100009), backups (SAP Note 1642148) or system replication (SAP Note 1999880) that prevent a release and re-use of existing redo log files
  • Preserving a high number of redo log files to make sure that there are re-usable files available (i.e. minimize RECLAIM LOG executions)
  • You can use the following parameter to make sure that the defined amount of redo log files is preformatted, so that no on-demand formatting is required: 
    indexserver.ini -> [persistence] -> log_preformat_segment_count = '<number_of_redo_logs>'

The related call stacks typically contain DataRecovery::LoggerImpl::reserveLogWriter and DataRecovery::LoggerImpl::reserveLog.
Mutex Wait *_fox_FoxFinalView_cache
*_lookup_lookup_cache
*_result_snapshot		 These named locks are used while populating cache nodes in planning sessions. The complete lock names include also variables like schema, planning session or calc scenario, e.g.:

SAPSR3:_SYS_PLE:20161110140816_6401020:DATA.cv034_fox_cv032_fox_FoxFinalView_cache
SAPSR3:_SYS_PLE:20161110140816_8775690:DATA.cv005_lookup_cv004_lookup_lookup_cache
SAPSR3:_SYS_PLE:20161110140816_6401020:DATA.00011_result_snapshot
Mutex Wait merger_queue_mutex This lock can show up when the result of an OLAP engine request, parallelized across several job workers, is merged. See SAP Note 1999997 (“Which general optimizations exist for reducing the SQL statement memory requirements?”) in order to make sure that SQL statements are executed as efficient as possible.
Mutex Wait Mutex[RowLockConflictManager.cpp:97]@0x<id>: TrexStore::SyncQueue::SyncQueue This lock is usually the consequence of other, general issues (e.g. memory, CPU, I/O, network or system replication). Therefore you should check for other symptoms you see at the time when these locks appear.

See SAP Note 2000000 for a general SAP HANA performance check.
Mutex Wait NameserverLock_ReadLock
NameserverLock_WriteLock		 See “How can special lock waits be analyzed?” -> “Nameserver topology” below for more information.
Mutex Wait PerformanceTracer_FileLock This lock is used when creating a performance trace. See SAP Note 1787489 and make sure that the performance trace is only activated when really required.
Mutex Wait PersistenceSessionRegistry This lock is required to open and close sessions on persistence level (SAP Note 2400005), e.g. during ptime::Connection::close or ptime::Connection::open). The related call stack modules are:

DataAccess::PersistenceSessionRegistry::registerSession
DataAccess::PersistenceSessionRegistry::unregisterSession

Long wait times can be caused by a thread that allocated this mutex and has to wait for “Barrier Wait” for the so called iteration barrier for quite some time. See the related section related to “Barrier Wait” above.
Mutex Wait PoolAllocator-MemoryPool Starting with SAP HANA 1.00.122.12 and 2.00.012.01 the former “unnamed Mutex” in MemoryManager::MemoryPool::allocate is explicitly named with PoolAllocator-MemoryPool.

This lock secures accesses to heap memory. Typically this wait is observed during memory defragmentation (SAP Note 1999997).
Mutex Wait Queue.cpp: queueserver2::Queue::Queue Idle wait of “Request” / “Queue Pull” threads from preprocessor to indexserver (see SAP Note 2114710), usually not critical
Mutex Wait Sequence Manager These waits are linked to bottlenecks when maintaining SAP HANA sequences. The related threads are typically working in modules ptime::SequenceManager::fillEnvBuffer / ptime::SequenceManager::nextval. Contention is possible if many tables with sequences are modified at the same time. In this case you can consider to activate or improve caching for frequently used sequences, e.g. with 100 elements:

ALTER SEQUENCE "<sequence_name>" CACHE 100

An overview of existing sequences can be retrieved via SQL: “HANA_Configuration_Sequences” (SAP Note 1969700).

Due to a bug in SAP HANA <= 1.00.122.11 it can happen that remote accesses to the CurrvalLocationMap happen in scale-out environments although the information should be available locally (module ptime::RemoteSequence::updateCurrvalLocationMap). Moving tables with critical sequences to the nodes with the CurrvalLocationMap can improve the accesses and optimize performance. Starting with SAP HANA Rev. 1.00.122.12 relevant map information is generally retrieved locally and the remote calls are no longer required.
Mutex Wait SessionMutex This mutex is used for different purposes in the areas of SqlExecutor thread handling and channel management. These mutex waits are typically symptoms of other underlying issues, so you should check for general problems at the time of the waits (e.g. memory reclaims, I/O or network issues) and have a look at the generated database traces (SAP Note 2380176). If required, open a SAP incident in order to request assistance from SAP support. Optimally you generate a runtime dump (SAP Note 2400007) at the time of the waits.
Mutex Wait system replication: SendQueueLock This lock secures the send queue in system replication environments (SAP Note 1999880) that contains shipping information for log buffers. It is reset in scenarios like disconnect or log shipping timeout. Waits are typically a secondary effect of issues in areas like system replication (SAP Note 1999880) or I/O (SAP Note 1999930).
Mutex Wait ThreadLockInfo This lock secures the ring buffer for thread samples (SAP Note 2114710). In general this lock is only held for a short time, but in rare cases a thread can be stuck while holding this lock:

  • Deadlock between parameter reconfiguration in ptime::ServiceThreadSamplesMonitor::reconfigure waiting for “Semaphore Wait” / “Event…: CallStack Not Initialized” and the running service thread sampler. In this case a SAP HANA restart is required. This bug is fixed with SAP HANA Rev. 1.00.122.12 and higher.
Mutex Wait TRexAPI_RefCountedLock_Lock This lock synchronizes trigger activities. Check if and why triggers or related features like sequences are implemented on the underlying tables and optimize or reduce them.
Mutex Wait TRexAPI_TableDeltaMerge_statisticLock This lock is held while CPU and memory information is collected. This information is required to determine a proper number of merge tokens. See SAP Note 2057046 for more information.
Mutex Wait TrexNet_Requestor_EndPointsLock Among others, the critical path of this lock contains domain name service (DNS) accesses (e.g. “Communication::getHostAddress”) up to SAP HANA SPS 09. If the DNS lookups are slowing down, threads may queue up on the TrexNet_Requestor_EndPointsLock lock. Therefore you should check if there are issues with DNS if you see threads queueing on this lock. See SAP Note 2222200 for more information regarding network checks in SAP HANA environments.

Consider switching the listeninterface parameter from .global to .internal in scale-out environments as recommended in SAP Note 2183363, because with .internal setting will suppress DNS requests.
Mutex Wait UndoAnchor This lock secures the materialization of undo data in parallelized update transactions. Contention is typically a side-effect of major underlying issues in areas like I/O (SAP Note 1999930) or system replication (SAP Note 1999880).
Mutex Wait unnamed Mutex Normally all important mutexes should be named properly, so “unnamed Mutex” should rarely pop up. If you face significant “unnamed Mutex” waits, consider the following known constellations:

  • While memory defragmentation (SAP Note 1999997) is active, threads may wait with “unnamed Mutex” in call stack MemoryManager::MemoryPool::allocate. The defragmentation thread (e.g. MemoryCompactor) is active in MemoryManager::freeSystemPagesInternal and System::UX::munmap. Starting with SAP HANA 1.00.122.12 and 2.00.012.01 the “unnamed Mutex” is explicitly named with PoolAllocator-MemoryPool.
  • If DML operations are executed and committed with a high frequency (e.g. because of concurrent single record INSERTs during migrations and data loads), the undo free list can become a central bottleneck, persistence garbage collection is no longer able to keep up and result in mutex waits of DML operations. The call stack then shows a location like “getPageWithFreeSpaceFromFreeList”. Avoid frequent COMMITs of small DML operations to come around this issue. See SAP Note 2169283 for more information related to SAP HANA garbage collection.
Mutex Wait ZipResultInput_SectionLock This lock can appear in the context of parallel sorting. Only threads of the same statement execution can block itself, so it is usually a local effect. If this lock appears frequently, you should analyze and optimize the related SQL statements (SAP Note 2000002).
Resource Load Wait This thread state is used when a resource requested from the resource container is currently loaded. In this situation the requesting thread needs to wait until the read I/O completes. If you see significant waits, the following reasons are possible:

  • I/O read bottlenecks (SAP Note 1999930)
  • Resource container operations called too often (SAP Note 2100040 -> “How can CPU intensive operations in SAP HANA be identified and optimized?” -> “PageAccess::LogicalPageAccessImpl::loadPageInternal”)
Mutex Wait

Semaphore Wait

Semaphore Wait

 unnamed Mutex

AsyncRequest.cpp: TrexNet::RequestSet::go

AsyncRequest.cpp: (anonymous namespace)::IOThread::run

 If you see these kinds of locks and only accesses to a certain table are blocked, you can check for the related thread types and thread methods (see SAP Note 2114710). In case you see significant amounts of

  • THREAD_TYPE: Request, THREAD_METHOD: __globalTransControl
  • THREAD_TYPE: AsyncWaiter, THREAD_METHOD: waiting

it is possible that the problem is caused by problems when modifying a partitioned table with more than one unique index. Check if the involved table has more than one unique index and is partitioned (across different nodes). If yes, check if it is possible to implement a workaround like disabling partitioning on this table. In particular you should avoid partitioning of BW SID tables (/BI0/S*). The underlying problem is fixed as of Rev. 97.
Semaphore Wait BackupDestBackint_Executor::Connect This lock happens when a SAP HANA backup waits for a 3rd party backup tool connecting to the Backint pipe. See SAP Note 1730932 for more information related to Backint.
Semaphore Wait BackupDest_DestCallbackSemaphore This lock secures the backup output channel (e.g. target file or backint pipe). It is normal to see this lock in these contexts. In case of very long durations you can check for problems in the I/O area. See SAP Note 1999930 for more information.
Semaphore Wait BackupSuperblockRead This lock secures the backup input channel. It is normal to see this lock in these contexts. In case of very long durations you can check for problems in the I/O area. See SAP Note 1999930 for more information.
Semaphore Wait capacityReached This is a normal wait situation during backups. The backup executor thread (Backup::BackupExe_Job) waits until all data is written to the backup destination.
Semaphore Wait ChannelUtilsSynchronousCopyHandler This lock is used when data is copied on disk level. This mainly happens during backups and system replication (transfer of replica data). It is normal to see this lock in these contexts. In case of very long durations you can check for problems in the I/O area. See SAP Note 1999930 for more information.

Related call stack modules are:

Stream::SynchronousCopyHandler::doCopy
Stream::ChannelUtils::copySynchronous
DataAccess::ReplicationProtocolPrimaryHandler::writeBackupToSecondary
Semaphore Wait CSV_ColumnExport_Semaphore_Stats This semaphore is required for synchronizing CSV exports. Due to a bug it can happen that parallel CSV threads infinitely wait for this semaphore and cancelling isn’t possible. This problem is fixed with SAP HANA Rev. >= 122.07, >= 2.00.002 and >= 2.00.010. As a workaround you can disable export parallelism:

indexserver.ini -> [import_export] -> use_parallel_export = false
Semaphore Wait DeltaIndexManager.cpp: TRexAPI::DeltaIndexManager::MergeAttributeThread::MergeAttributeThread This lock can show up when the MergedogMonitor thread waits for merges performed by the MergeAttributeThread. See SAP Note 2057046 for more information related to delta merges.
Semaphore Wait Event …: CallStack Not Initialized This is a generic, unspecific semaphore name used for various semaphores in IBM on Power environments. On Intel they follow the naming convention “<module>.cpp: <function>” instead, e.g. “DeltaIndexManager.cpp: TRexAPI::DeltaIndexManager::MergeAttributeThread::MergeAttributeThread”.
Semaphore Wait

Mutex Wait

ExclusiveLock Enter

 IndexMgr.cpp: TRexConfig::IndexHandle::acquire
Acquire Index Access

unnamed Mutex

TRexConfig_IndexMgrIndex_BucketLock
TRexConfig_IndexMgrIndex_FastLock
TRexConfig_IndexMgrIndex_Lock
unnamed Mutex

 These waits are linked to the IndexHandle, a column store table level synchronization lock that is mainly used for delta merge (SAP Note 2057046) synchronization.

Typical reasons for these wait situations are:

Situation Details
High merge frequency As these locks are mainly linked to merges, an unnecessary high amount of delta merges can intensify the problem. See SAP Note 2057046 and make sure that the merge and optimize compression configuration is in line with the SAP HANA parameter recommendations (SAP Note 2186744).
Move of uncommitted rows from current to new delta storage During a delta merge uncommitted rows need to be moved from the source delta storage to the target delta storage. When this happens, a lock is set on the table and change operations are blocked. See SAP Notes 2098801 and 2057046 for more information and make sure that the merge decision functions are set properly.
Lock escalation by delta merge If another transaction holds a – usually non-critical – shared lock and the delta merge wants to move uncommitted rows to the new delta storage, it escalates the lock to a critical exclusive lock blocking all kind of operations. This can e.g. happen in the following situations:

  • If on Rev. <= 102 both the table changes and the delta merge are blocked by this lock and at the same time the involved table has a LOB field that is frequently changed, the lock holder can be the LobGarbageCollector (see SAP Note 2169283 regarding LOB garbage collection). Due to a bug the LOB garbage collector can run much longer than necessary and at the same time the auto merge escalates the usual shared lock to an exclusive lock. As a workaround you can disable auto merges on the underlying table and trigger hard merges during non-critical times instead (see SAP Note 2057046). As a permanent solution implement Revision 102.01 or higher.
  • With Rev. <= 102.05 and 110 to 112.01 an UPSERT with a sub-SELECT on a table with trigger can take quite a long time (call stacks containing TRexAPI::LateMatColumnSource::getValue) and the lock can be escalated by a concurrent delta merge on the same table (SAP Note 2327289)

If you suffer from this situation on a specific table, you can disable auto merge as a workaround and perform a hard merge during non-critical times.

Starting with SAP HANA 2.0 the index manager is redesigned and the probability of lock waits is reduced.
Expensive queries Expensive queries can allocate the IndexHandle lock in specific phases for a long time, e. g.:

  • Writing data to an internal table (call stack: ItabCreateGuard::~ItabCreateGuard).
  • Accessing column information during a full table scan (call stack: *::mgetSearch, e.g. UnifiedTable::impl::VersionedNBitVectorDataObject::mgetSearch), See SAP Notes 2318281 and 2321573
  • Evaluation of filter condition (call stack: qo3::OptimizerImpl::evaluate)
  • Expensive sorting (call stack: TRexUtils::Parallel::Sorter; thread method / detail: Sorter / RangeSortJob)
  • Single threaded OLTP search (call stack: Executor::PlanExecutor::doCalculateSingleThreaded)

The best solution in this case is to tune these bad queries, so that both their own runtime and the effect on the overall system is minimized. See SAP Note 2000002 for more information related to SQL statement tuning.

In order to release a currently existing lock, you can terminate the related session as described in SAP Note 2092196.
Export Exporting a table also requires the IndexHandle. As a consequence accesses to a table, that is currently exported, can be stuck acquiring the IndexHandle. Therefore you should avoid exporting large tables in parallel to critical business activities.
Column load During SAP HANA column loads (see SAP Note 2127458) an IndexHandle of the related table is held. Under normal situations this is no issue, because the table can anyway not be accessed unless it is properly loaded. In some scenarios it can be a problem (e.g. if you want to drop a large, faulty table and the DROP doesn’t succeed due to a long load operation). In this case you have to make sure that no load happens (e.g. by disabling indexserver.ini -> [sql] -> reload_tables temporarily and by avoiding accesses that would load the table).

Up to SAP HANA SPS 11 it is often not easily possible to identify the IndexHandle lock holder. Starting with SAP HANA SPS 12 you can use the following hdbcons functionality to determine IndexHandle wait situations:

hdbcons 'indexmanager waitgraph -w'

The overall IndexHandle state can be determined with the following command:

hdbcons 'indexmanager list [-f <table>]'

See SAP Note 2222218 for more information related to hdbcons.

Starting with SAP HANA SPS 12 also runtime dumps (SAP Note 2400007) contain a new section [INDEXMANAGER_WAITGRAPH] with lock dependency information related to the IndexHandle.
Semaphore Wait IpmmTaskWait
MemoryReclaim		 These inter process memory management (IPMM) locks can show up when memory reclaim activities like self compactions or shrinks happen. See SAP Note 1999997 and make sure that memory is used efficiently so that the amount of reclaim operations is reduced.

Threads waiting for this lock typically work in module MemoryManager::GlobalMemoryHandler::pmAcquireIPMMLock.

In the [IPMM_MEMORY] of the runtime dumps (SAP Note 2400007) you are able to identify the process holding the lock:

[6] PID=16290 (state=T), SId=1547305659, compactors active, alive, process name: hdbnsutil
    AB=1228591104b (1.14gb), UA=0b, U=109622220b (104.54mb), FSL=0b, SMAS=1191825408b (1.10gb), ...
    ! Executing shrink. Process with SlotID 5 requests 0b to shrink, 
    13146370048b (12.24gb) to reserve, task: DefragmentationTask.

Up to SAP HANA SPS 10 also the hdbnsutil tool is part of IPMM. If hdbnsutil holds the lock, you should check if this process is stuck for some reason and terminate it manually. Starting with SAP HANA SPS 11 hdbnsutil is no longer part of IPMM and so it can no longer be responsible for the IPMM locks.

Shrink activities are documented in the database trace (SAP Note 2380176) with entries like “Information about shrink”. Self compactions (defragmentations) aren’t displayed in the trace files. You can find details via
Semaphore Wait LogBufferFreeWait This lock indicates that no freespace is available in the log buffer to record redo information. This is typically caused by disk I/O issues. See SAP Note 1999930 for more information.
Semaphore Wait MailSenderCallback Waits for this semaphore can be caused by the statistics server having problem to send out mails. In the related statistics server trace file you may find an entry related to the SMTP port 25 like:

(invalid)-><ip_address>/25_tcp ConnectWait

The permanent hanging situation is caused by SAP HANA bugs which are fixed with Rev. 83 and Rev. 90.

As a consequence of this problem the statistics server may no longer be replicated properly. In order to restart the system replication, you can restart the statistics server.
Semaphore Wait PageIO::SyncCallbackSemaphore Starting with SAP HANA 2.0 synchronous I/O requests are implicitly executed as asynchronous I/O requests. The PageIO::SyncCallbackSemaphore indicates that the thread is currently waiting for the asynchronous I/O request to finish properly. See SAP Note 1999930 for tuning SAP HANA I/O performance.
Semaphore Wait POSTCOMMIT_FINISH_SMP This lock is related to COMMIT processing and can be caused by the following issues:

  • Problems with system replication like increased log shipping latencies or full log shipping buffer. See SAP Note 1999880 -> “Can problems with system replication impact the performance on the primary system?” for more details.
  • I/O bottlenecks on primary system (see SAP Note 1999930)
  • Other general issues on primary system
Semaphore Wait PrefetchCallback
PrefetchIteratorCallback		 These waits are active when the thread waits for pages requested from disk. A high number of these waits can indicate that the I/O stack is not able to keep up with the number of I/O requests. This may be normal (e.g. when a large number of columns is loaded in parallel after startup), but it can also indicate bottlenecks in the I/O area. See SAP Note 1999930 for more information regarding I/O optimization.

Threads waiting for this lock typically have call stacks including modules like DataContainer::PageChainContainerPrefetchedIterator::PrefetchBuffer::getNextPage.
Semaphore Wait PrivateLogBufferFlushWaitSemaphore This semaphore is required while flushing session specific log buffer information into the global log buffer in memory. It can show up if there is contention on the global log buffer, e.g. due to LoggerBufferSwitch waits. Therefore you should check for other unusual lock situations happening at the same time and optimize these lock situations.

Private log buffers are used as of SAP HANA 1.00 SPS 10.
Semaphore Wait ResourceFlushWaitA
ResourceFlushWaitB		 These locks indicate that a SAP HANA thread (e.g. in the context of a savepoint) waits for asynchronous I/O write requests to be finished. Long lock durations can indicate a high write volume or problems in the I/O area. See SAP Note 1999930 for more information regarding I/O optimization.
Semaphore Wait SaveMergedAttributeJobSemaphore This lock is used when results of a column thread should be merged to disk but the actual merge hasn’t finished, yet. See SAP Note 2057046 for more information related to merges.

You can normally consider this lock as a non-critical idle wait.
Semaphore Wait TableDeltaMerge.cpp: TRexAPI::TableMergeRequest::TableMergeRequest Threads waiting in modules like TRexAPI::MergeMonitor::acquireToken for this lock, the number of merge tokens is too small to handle the current merge load. Occasional merge peaks are acceptable, but if it happens frequently, you should consider the following optimizations:

  • Check if the amount of DML change operations can be reduced, so that the merge pressure is implicitly reduced.
  • Check if the merges performance can be improved.
  • Check if there are merge failures on large tables (e.g. due to OOM).
  • Increase the number of merge tokens if feasible.

See SAP Note 2057046 for more details about merges and merge tokens.
Semaphore Wait TableLoad.cpp: TRexAPI::LoadFieldThread
TableLoad.cpp: ltt::vector::vector		 This semaphore is used when TableReload threads trigger LoadField threads in order to load columns into memory, e.g. after a startup of SAP HANA. See SAP Note 2127458 for more information related to column loads. It is normal to see spikes of this wait at some times. Its runtime mainly depends on the I/O performance when columns are read from the data area on disk. If it takes too long, you should check the I/O performance based on SAP Note 1999930.
Semaphore Wait

SharedLock Enter
ExclusiveLock Enter

 TranslationTable.cpp: JoinEvaluator::TranslationTable::fill
TranslationTable.cpp: JoinEvaluator::TranslationTable::completeFill

RWLock[TranslationTable.cpp]: JoinEvaluator::TranslationTable::TranslationTable

 These locks happen in the context of populating translation tables which are used to map join columns (see SAP Note 1998599). Check the following suggestions to optimize the situation:

  • See SAP Note 2124112 and check if there are SQL statements with a high number of instantiations in the SQL cache as – in case of joins – they are candidates for creating a significant amount of translation tables.
  • See SAP Note 2217936 and make sure that caching of translation tables for temporary tables is deactivated.
  • If this is not possible (e.g. because you are on a release <= 97.01), you can consider increasing the translation table cache size (indexserver.ini -> [joins] -> translator_cache_size). You can for example start with doubling the existing value. Be aware that this setting is a trade-off between performance and memory consumption, because the heap allocator Pool/JoinEvaluator/TranslationTable will increase (SAP Note 1998599).
Semaphore Wait TRexApiSystem.cpp: TRexAPI::TRexApiSystem::preloadTables This wait is linked to column preload as part of SAP HANA startup (SAP Note 2127458). The related thread is of type ‘Assign’ with  thread method ‘post_assign’. In case of longer waits you have to check why preload takes longer than expected. Typical reasons are individual issues (e.g. long text column load as described in SAP Note 2461411 or I/O read bottlenecks (SAP Note 1999930).
Semaphore Wait unnamed Semaphore Normally all important semaphores should be named properly, so “unnamed Semaphore” should rarely pop up. If you face significant “unnamed Semaphore” waits, consider the following known constellations:

  • Persistence garbage collector threads (GCJob*, SAP Note 2169283) may block each other with “unnamed Semaphore” in call stack module DataContainer::VirtualFileStatsProxy::removeVirtualFileLOBStatistics or DataContainer::FileIDMapping::removeMapping. This has no direct effect on the business activities, but indirectly it can slow down parallelized database requests because it reduces the amount of available JobWorker threads limited by max_concurrency (SAP Note 2222250). With SAP HANA >= 2.0 this semaphore is properly named.
Semaphore Wait WaitAndSwitchCounterResourceSemaphore This lock is required when a savepoint or snapshot waits for the resource flush thread to finish flushing data pages to disk. It is normal that this wait appears from time to time. If you see a particularly high number of occurrences, you can check the savepoint behavior as described in SAP Note 2100009.
SharedLock Enter
ExclusiveLockEnter		 Acquire AttributeStore Access This lock secures access to the column store for specific operations. For example, a compression optimization (SAP Note 2112604) has to access the column store without any garbage collection (SAP Note 2169283) going on in parallel (AttributeEngine::AttributeStoreHandle::setStoreHandleAndLockGC), and so it has to wait for ‘Acquire AttributeStore Access’ until a running persistence garbage collection is finished.
SharedLock Enter
ExclusiveLock Enter

Mutex Wait

 AttributeValueContainer writeLock

AttributeValueContainer readLock

 These locks secure accesses to column store main areas:

  • “AttributeValueContainer readLock” secures the current record version.
  • “AttributeValueContainer writeLock” secures the next record version during delta merge (SAP Note 2057046).

Lock contention on “AttributeValueContainer readLock” can happen in the following cases:

  • It is held while a column is loaded into memory, so you should check if large columns are unnecessarily unloaded and reloaded (SAP Note 2127458), e.g. due to memory shortage (SAP Note 1999997). Also an inverted index load after optimize compression can result in these lock situations (SAP Note 2127458 -> “When do loads happen?”).
  • A delta merge (SAP Note  2057046) holds an exclusive read lock in some phases, e.g. when writing down the table data to disk (module AttributeEngine::MemoryAvc2::save).
  • An exclusive lock is also held during import operations.
SharedLock Enter
ExclusiveLock Enter		 calcEngine_CalcEngineManager_ScenarioMapRWLock These locks can be secondary symptoms when a thread is blocked by the ceManager_GCLock lock, i.e. by the calculation engine garbage collector (CalcEngineGarbageCollector, see SAP Note 2169283).
 SharedLock Enter
ExclusiveLock Enter		 Connection List RWLock This lock secures the session connection list. With SAP HANA Rev. 110 to 112.02 it can be critical in the context of the SAP HANA bug described in SAP Note 2400515.
SharedLock Enter
ExclusiveLock Enter		 ConsistentChangeLock During the blocking phase of savepoints DML operations are blocked with “SharedLock Enter” related to the lock type ConsistentChangeLock. The call stacks contain modules like DataAccess::SavepointSPI::lockSavepoint. See SAP Note 2100009 and check if the blocking phase of savepoints can be optimized.

The actual reason of long ConsistentChangeLock waits can also originate from other activities. The section [SAVEPOINT_SHAREDLOCK_OWNERS] of runtime dumps (SAP Note 2400007) lists owners of shared ConsistentChangeLocks. If this information isn’t available, you can double-check the following list of potential lock holders:

  • With SAP HANA SPS 09 and 10 the savepoint can be blocked with a ConsistentChangeLock by the MergedogMonitor thread (SAP Note 2057046) being active for a long time in modules like TRexUtils::BitVector::beginAt, TrexStore::UdivsMVCC::deleteFromBaseList, TrexStore::UdivListContainerMVCC::adaptUdivsAfterMerge or TrexStore::UdivListManager::adaptValidUDIVs. This situation typically happens after having deleted a large amount of records in a table. Starting with SAP HANA SPS 11 this critical merge phase is improved from a performance perspective. On earlier Revisions you can check which table is responsible for the long merge activities and take appropriate actions to improve merge times (e.g. reduction of changes, partitioning, adjustment of merge parameters).
  • With SAP HANA <= 102.05 and SAP HANA 110 – 112.01 it can happen that a LocalPostCommitHandler holds a ConsistentChangeLock while cleaning up record locks (typical call stack modules: TrexStore::RowLockManager2::clearLocks, TrexStore::UdivListContainer::clearLocks2, TrexStore::UdivListManager::releaseRowLocks). This issue can happen if a particularly high amount of records is modified by a transaction. Technically this situation can be avoided by locking the whole table for modifications before performing the modifications (LOCK TABLE “<table>” IN EXCLUSIVE MODE). This issue is fixed as of SAP HANA Rev. 102.06 and 112.02.
  • With SAP HANA 90 to 112.01 deadlocks involving ConsistentChangeLock waits can be caused by delta migrations after an upgrade to SAP HANA >= SPS 09 (SAP Note 2304318). The call stack module DataContainer::PersistentObjectContainerImpl::migratePersistentObjects is an indication for this kind of delta migration.
  • With SAP HANA Rev. <= 122.04 the delta merge (SAP Note 2057046) of tables with LOB columns keeps the ConsistentChangeLock for an unnecessary long time. This behavior is improved starting with SAP HANA Rev. 122.05.
  • With SAP HANA <= 122.04 an ALTER TABLE … PARTITION can hold the lock for a significant time if the amount of records in a partition is already close to 2 billion (SAP Note 2399151).
  • With SAP HANA <= 122.05 the SegmentPreallocator thread can hold the ConsistentChangeLock for a long time while itself waiting for the segment allocation latch (module ptime::SegMgr::allocate). The segment allocation latch can be held a significant time by queries on the internal monitoring view M_DEV_RS_PAGES (that may be accessed in context of analyzing the issues described in SAP Notes 2083100 and 2362759).
  • SAP Note 2404252 describes a deadlock between ConsistentChangeLock and GlobalContainerNameDirectoryTransLockStats in the context of M_TABLE_PERSISTENCE_LOCATIONS accesses on SAP HANA Rev. 122.01 to 122.05.
  • Time synchronization (modules like ptime::Transaction::setPrecommitTs, Newdb::GlobalTransClientSync::sendTimestamp or ptime::MasterTimestampBroadcaster::execute) can hold the ConsistentChangeLock with SAP HANA Rev. <= 122.01. This problem is fixed starting with SAP HANA Rev. 122.02.
  • When table optimizations (SAP Note 2057046) write non-string data down to disk (module AttributeEngine::MemoryAvc2::save), they hold the ConsistentChangeLock. A performance improvement is available starting with Rev. 1.00.122.07 and 2.00.002.
  • The ConsistentChangeLock can also be held by threads waiting for FileIDMapping locks, so ConsistentChangeLock waits can be a secondary symptom of FileIDMapping contention.
  • Table optimizations (SAP Note 2057046) in steps like UnifiedTable::MVCCWriter::reorderMVCCData, TrexStore::UdivListContainerMVCC::reorderUdivs or TrexStore::UdivListManager::reorderUdivs also hold the ConsistentChangeLock. This scenario is optimized with SAP HANA Rev. >= 122.09, 2.00.002.01 and 2.00.010.
  • An increased log switch wait count ratio (SAP Note 2215131) can be responsible for LoggerBufferSwitch and LogBufferFreeWait waits that hold the ConsistentChangeLock and may block the waitForLock phase of savepoints. Optimize the log switch wait count ratio in order to minimize the waitForLock savepoint phase in this context.
SharedLock Enter
ExclusiveLock Enter		 ExpensiveStatements_RingBufferLock This lock secures the ring buffer linked to the expensive statement trace (SAP Note 2180165). Typical reasons for an increased amount of lock waits are:

  • General SAP HANA performance issue (SAP Note 2000000) that slows down a lot of database requests, so that more requests than usual are traced
  • Rather low trace threshold values (e.g. global.ini -> [expensive_statement] -> threshold_duration) leading to a high amount of traced requests; check ID 623 (“Traced expensive SQL statements (last day)”) of SAP HANA Mini Checks (SAP Note 1999993) can indicate a low threshold setting in this context
SharedLock Enter
ExclusiveLock Enter		 FileIDMapping This lock secures persistence container accesses (e.g. for hybrid LOBs or virtual files). A high contention can happen when many persistence garbage collector threads (GCJob*) are concurrently active. Typical call stacks are DataContainer::FileIDMapping::addMapping (e.g. in the context of an INSERT involving a disk LOB), DataContainer::FileIDMapping::mapToEntryID and DataContainer::ContainerDirectoryImpl::getInternal. In this case you can reduce the maximum number of persistence garbage collectors by setting the following parameter to a lower value (SAP Note 2222250).

global.ini -> [persistence] -> max_gc_parallelity

See SAP Note 2169283 for more information regarding SAP HANA garbage collection.
SharedLock Enter
ExclusiveLock Enter		 GeometryDeltaAttribute Lock This lock is used to synchronize accesses to data structures on the delta of geometry columns (ST_POINT and ST_GEOMETRY) in the context of modification operations like INSERT or UPDATE.
SharedLock Enter
ExclusiveLock Enter		 GlobalContainerNameDirectoryTransLockStats This lock protects the persistence container name directory. If concurrent garbage collector threads (GCJob*) are responsible for the lock waits you can reduce the maximum number of persistence garbage collectors by setting the following parameter to a lower value (SAP Note 2222250).

global.ini -> [persistence] -> max_gc_parallelity

See SAP Note 2169283 for more information regarding SAP HANA garbage collection.

SAP Note 2404252 describes a SAP HANA bug with Rev. 122.01 to 122.05 that can result in lock waits in combination of M_TABLE_PERSISTENCE_LOCATIONS accesses.
SharedLock Enter
ExclusiveLock Enter		 Global SqlExecutor thread lock This lock is used when the number of SqlExecutor threads is changed for some reason, e.g.:

  • Adjustment of number of available SqlExecutor threads via parameter <service>.ini -> [sql] -> sql_executors (SAP Note 2222250)
  • Delegation of work to another SqlExecutor thread in case of a lock wait situation
  • Implicit creation of additional SqlExecutor thread in context of <service>.ini -> [session] -> busy_executor_threshold (SAP Note 2222250)

Waits for this lock are typically a consequence of another underlying root cause (e.g. very high load or infrastructure bottlenecks).
SharedLock Enter
ExclusiveLock Enter		 IndexManagerMapLock This lock secures table accesses on a low level. Usually these locks are shared and so no waits happen. In specific scenarios locking can happen, usually in combination with DDL commands or background  activities (e.g. merge, optimize compression). This lock is global and not table-specific, so accesses to all tables can be hampered. Check the recommendations related to the following locks to identify the root cause of IndexManagerMapLock issues:

  • Acquire Index Access
  • BTree GuardContainer
SharedLock Enter
ExclusiveLock Enter		 JoinEvaluator_JEPlanData_Lock This lock situation happens when multiple threads of the same join engine job need to access the same plan data objects. This is a rather normal situation so that it is not necessarily an issue that this lock belongs to the most frequent ones in a system. The lock waits are always local to a single request / single connection. If the locks result in performance issues, you should check if the underlying SQL statement is already processed optimally or if it can be tuned (SAP Note 2000002).
SharedLock Enter
ExclusiveLock Enter		 LlangCompilingMap This lock synchronizes concurrent compilation of L procedures.
SharedLock Enter
ExclusiveLock Enter		 MergeMonitor_Lock
TRexAPI::TableMergeRequest::TableMergeRequest		 These lock wait situations indicate that merge operations have to wait for merge tokens, e.g. due to a high amount of concurrent merges or bottlenecks during merging. See SAP Note 2057046 for more details. In very special cases you can adjust parameters like the following in order to control the merge throughput:

indexserver.ini -> [mergedog] -> load_balancing_func
indexserver.ini -> [mergedog] -> token_per_table
SharedLock Enter
ExclusiveLock Enter		 MergeMVCCLock This lock can happen in case of concurrent merges and is typically a consequence of other issues (e.g. I/O problems or highly concurrent merges). See SAP Note 2057046 for more details regarding merges.
SharedLock Enter
ExclusiveLock Enter		 MVCCSyncLock This lock is relevant for column store tables only and it is responsible for synchronizing low-level MVCC operations with DDL statements, delta merges and compression optimizations. It can e.g. show up if different internal activities like persistence garbage collection and compression optimization collide. See SAP Note 2169283 for more information related to garbage collection and MVCC.
SharedLock Enter
ExclusiveLock Enter		 pageAllocLatch This latch secures shared memory page allocations in the context of the row store. Contention for this latch was observed in the context of MemoryReclaim waits (SAP Note 1999997). If you face a high amount of pageAllocLatch waits, you should check at first if there was a concurrent shrink operation and MemoryReclaim waits. If yes, the pageAllocLatch is most likely only a secondary symptom and you have to reduce the risk of automatic memory shrinks.
SharedLock Enter
ExclusiveLock Enter		 SavepointLock This lock synchronizes savepoints and makes sure that a savepoint (or snapshot) isn’t started at a time when another savepoint (or snapshot) is already running. As an example, a system replication related snapshot will wait for SavepointLock as long as a normal savepoint is in progress on primary side. If you frequently see these locks, you should check and optimize the savepoint runtime and performance (SAP Note 2100009).
SharedLock Enter
ExclusiveLock Enter		 SessionContextCacheMapLock This lock secures the access to session context information. If you see many waits, it is usually not a problem itself, but a symptom of another underlying issue (e.g. slow or hanging COMMIT operations).
SharedLock Enter
ExclusiveLock Enter		 transLock This lock secures transaction management. Permanent transLock waits can be caused by a SAP HANA deadlock bug that is fixed with Rev. 112.04 and Rev. 121. If you already run into this bug situation, you have to restart SAP HANA to resolve the stuck situation.

If the waiting threads are in a call stack containing ptime::Transaction::find_trans_from_tid2trans and ptime::TransTokenConnector::checkTIDGloballyClosed, the problem is linked to NO LOGGING tables and you could prevent future issues by making sure that all of these tables are located on the master node. See SAP Note 2340583 for more information.
SharedLock Enter
ExclusiveLock Enter		 TRexConfig_IndexMgrIndex_ReplayLock This lock is required when delta log information is loaded or replayed. If for example a large delta storage isn’t loaded into memory, yet, the first thread requiring this information will load it in a module like Delta::log::Replay::replayLog. Other processes have to wait for the TRexConfig_IndexMgrIndex_ReplayLock lock.

Typical reasons for this lock are:

  • Load of many large table (partitions) after startup and rather small value of parameter for indexserver.ini -> [parallel] -> tables_preloaded_in_parallel (see SAP Note 2222250); increasing the parameter may help to reduce the contention.
  • Disk I/O bottlenecks when reading columns (SAP Note 1999930)
  • In some situations this situation can also happen when delta merges fail due to load issues. You can check according to SAP Note 2057046 if delta merges are running fine.
SharedLock Enter
ExclusiveLock Enter		 unnamed ReadWriteLock In some cases read write locks aren’t named properly and so they are only shown as “unnamed ReadWriteLock”. The most typical scenarios are:

Scenario Details
Thread method: GCJob …

DataContainer::FileIDMapping::removeMapping

 In this case persistence garbage collectors (SAP Note 2169283) block each other in the context of persistence container management. This has no direct effect on the business activities, but indirectly it can slow down parallelized database requests because it reduces the amount of available JobWorker threads limited by max_concurrency (SAP Note 2222250).

With SAP HANA >= 2.0 the related locks are properly named.
SharedLock Enter
ExclusiveLock Enter

Semaphore Wait

 vfLOBStatsMapLock

unnamed Semaphore

 The vfLOBStatsMapLock is linked to virtual file LOB statistics. If these locks appear on secondary system replication sites (SAP Note 1999880) with operation mode logreplay in context of LogRecoveryQueue<id> threads and call stacks like DataContainer::VirtualFileStatsProxy::getVirtualFileLOBStatistics or DataContainer::VirtualFileStatsProxy::removeVirtualFileLOBStatistics, a reduction of the recovery queue size can help to improve the situation:

global.ini -> [persistence] -> recovery_queue_count

See SAP Note 2222250 for more information.

An overview of internal lock waits can be retrieved via:

  • M_READWRITELOCKS, M_SEMAPHORES, M_MUTEXES
  • SQL: “HANA_Locks_Internal_LockWaits_Overview” (SAP Note 1969700)
  • liveCache related internal lock waits can be checked via SQL: “HANA_liveCache_LockWaits_Current” (SAP Note 1969700)

Be aware that many harmless internal lock waits are included in these tables (e.g. threads being idle, but waiting for internal locks at that time), so a top down analysis is not helpful.

If the SAP HANA database is no longer accessible due to the internal lock escalations, see SAP Note 1999020 for potential additional analysis steps.

If you are not able to determine the root cause of critical internal lock waits, open a SAP incident on component HAN-DB and provide a full system dump if possible (SAP Note 1732157).

7. How can special lock waits be analyzed?

Special lock waits are all lock waits that aren’t covered by normal thread states and monitoring views. Due to a lack of monitoring data it is particularly hard to analyze them. In the following table you can find pro-active suggestions and further details:

Lock type Details
Nameserver topology The namesever topology tracks the current state of the database (e.g. existing tables and locations) and is implemented as shared memory. Concurrent accesses to the nameserver topology are synchronized via a file on operating system level:

SAP HANA SPS File
<= 09 /tmp/.hdb_<sid>_<inst_id>_lock
>= 10 /var/lib/hdb/<sid>/.hdb_<sid>_<inst_id>_lock

If there is a bottleneck accessing the file (e.g. because the /tmp file system is mounted via NFS), the performance can be significantly impacted. In order to minimize the risk of synchronization delays you have to make sure that the file is located on fast, local devices. See SAP Note 2100296 for more information. Starting with SAP HANA Rev. 91, but only with SPS 09, a message is written to the daemon trace if the file system is potentially critical. Mini check 1720 (“Supported nameserver lock file location”) evaluates this information. The different file type used with SPS >= 10 avoids these issues.

An increased number of nameserver topology accesses can also be linked to an inefficient SQL cache utilization, e.g. due to many evictions or a lack of bind variables. See SAP Note 2124112 and make sure that SQL cache and parsing activities are working fine.

In rare cases it can be helpful to relocate this file to a different location. This can be done by configuring the following parameter in configuration file /usr/sap/<sid>/HDB<inst_id>/<hostname>/sapprofile.ini:

HDB/NameServer/RWLockPath=<filesystem_path>

Example:

HDB/NameServer/RWLockPath=/var/hdb/ramdisk

Typical call stacks of threads waiting for the lock are:

TrexIpc::NameserverLock::acquireRead
NameServer::FTreeContainer::getValue

An indirect indication for problems accessing this file can be increased SelfWatchDog thread activities (see SAP Note 2114710). Mini Check 870 performs a related check (see SAP Note 1999993). If you assume a problem, you can run SQL: “HANA_Threads_Callstacks” (THREAD_TYPE = ‘SelfWatchDog’) available via SAP Note 1969700 in order to have a look at the call stacks of the SelfWatchDog threads. If threads typically work in the following call stack, a problem accessing the lock file is likely:

checkLockFile
__close_nocancel
__fopen_internal
__GI__IO_file_open
_IO_new_fclose
_IO_new_file_close_it
_IO_new_file_fopen
__open_nocancel
TrexService::SelfWatchdog::checkLockFile
TrexService::SelfWatchdog::run

Be aware that the following warning in trace files can actually be ignored because all file system types can be used for the lock file (unlike the DATA and LOG area where only specific file system types are supported):

w FileIO Filesystem.cpp: Unsupported file system "<file_system_type>" for "/var/lib/hdb/<sid>/.hdb_<sid>_<inst_id>_lock"
Runtime dump lock If several runtime dumps (including variations like OOM or crash dump) are triggered at the same time, only one dump is executed at a time and other have to wait in call stacks like:

1: __lll_lock_wait+0x20
2: _L_lock_1008+0xb
3: __pthread_mutex_lock_internal
4: Synchronization::SystemMutex::lock
5: Diagnose::CrashDump::executeRuntimeDump
6: Diagnose::RuntimeDump::createSystemDump
7: MemoryManager::reportMemoryLimitViolation


For technical reason the lock is implemented as system mutex (Synchronization::SystemMutex::lock) that is not displayed as SAP HANA lock wait. Instead the thread state appears as ‘Running’. At the same time it is an idle wait and no CPU is consumed.


A hanging runtime dump can – depending on the context of the dump – be responsible for contention on other internal SAP HANA locks. To avoid this kind of locking you should reduce the possibility that different types of runtime dumps are executed in parallel.











8. Do timeouts for lock waits exist?


Transactional lock waits are terminated when the time limit defined with the following parameter is exceeded:


indexserver.ini -> [transaction] -> lock_wait_timeout


Its default value 1800000 which represents 1800000 ms / 1800 s / 30 minutes. This means that a lock wait is terminated after 30 minutes and the following error message is issued:


SQL error 131: transaction rolled back by lock wait timeout


This behavior is different from other databases like Oracle where no timeout for exclusive lock waits exist and transactions will wait for the lock until the lock is available or the transaction is manually terminated.


For internal lock waits no timeout is implemented.


9. What happens in case of deadlocks?


Deadlocks are situations where two or more transactions lock each other cross-wise so that no transaction will ever be able to proceed. Usually deadlocks are caused by the application design, but there can be also more technical deadlocks in the context of primary key constraints (SAP Note 2429521). The actual behavior in case of a deadlock depends on the lock type:










Lock type
Deadlock behavior




Transactional lock
If SAP HANA recognizes a deadlock based on a transactional lock, one of the transaction will be terminated and the following error is issued:


SQL error 133: transaction rolled back by detected deadlock


This behavior is different from other databases like Oracle where a deadlock will only terminate the current DML operation and not the whole transaction.





Internal lock
A deadlock on top of internal locks (e.g. SAP Note 2304201) has to be considered as SAP HANA bug and can only be resolved with a restart.


You can check for certain (but not all) internal deadlocks using the hdbcons deadlock detector functionality as described in SAP Note 2222218:


hdbcons 'deadlockdetector wg -w -o <file_name>.dot'












10. Is it possible to identify the record responsible for a record lock?


It is currently not possible to identify the record responsible for a record lock. The view M_BLOCKED_TRANSACTIONS contains a reference to the record (WAITING_RECORD_ID), but this reference can’t be evaluated in a reasonable way.


11. How can I determine how long certain types of locks are held?


Normally it is most important to analyze lock wait situations, i.e. the concurrent access to the same resources. For transactional locks you can additionally see how long locks are held (even if no transaction is waiting) by using the following approaches:


    

  • M_OBJECT_LOCKS, M_RECORD_LOCKS, M_OBJECT_LOCK_STATISTICS
    • 

  • SQL: “HANA_Locks_Transactional_Current” / SQL: “HANA_Locks_Transactional_Total” (SAP Note 1969700)
    • 



12. What is the unit of column TOTAL_LOCK_WAIT_DURATION in monitoring view M_SQL_PLAN_CACHE?


TOTAL_LOCK_WAIT_DURATION contains the time a SQL statement had to wait for transactional locks. Like all other duration columns in monitoring view M_SQL_PLAN_CACHE the unit is usually microseconds. Due to a bug it changed to milliseconds for SAP HANA Revisions 110 to 112.06 an 120 to 122.03. As a consequence the displayed wait durations are factor 1000 too low on these Revision levels.


13. How can I automatically capture additional information in case of unusual locking issues?


In order to optimally understand specific – mainly internal – lock waits, additional information like runtime dumps (SAP Note 2400007) are helpful. You can schedule SAP HANASitter (SAP Note 2399979) in order to automatically trigger follow-up actions in case of unusual thread states.


Example:










Command
Details




python hanasitter.py … -ct LOCK_WAIT_NAME,Resource-Load-Wait,10
Trigger data collection if there are at least 10 threads waiting for the “Resource Load Wait” lock










14. Why are there deadlocks although changes are done in a sorted manner?


Deadlocks – i.e. terminations with “SQL error 133: transaction rolled back by detected deadlock” – are a consequence of cross-wise locks. If all concurrent transactions perform the record changes in the same order, no cross-wise lock can happen, so changes with a special sort order typically avoid deadlocks. Instead you can only see transactional lock waits.


There is one important restriction to this rule: If the underlying table is partitioned, the modifications are no longer necessarily done in the defined sort order, because modifications can happen in parallel in different transactions. This is a general risk when you partition a table, even if the partition columns and the WHERE clause of the modification match.


If you want to make sure that concurrent sorted modifications on a partitioned table don’t run into a deadlock, you have to set the partitioning related minimum bulk load size to a sufficiently high value in order to make sure that loads won’t use the bulk load parallelism, e.g.:


indexserver.ini -> [partitioning] -> bulk_load_size = 2000000000


This global setting may cause increased runtime of all modification operations on all partitioned tables in the system. Therefore it has to be used with care and tested thoroughly.


15. How can the lock holder be determined?


In case of wait situations it is of central importance to identify the lock holder and analyze its behavior. The following options exist to identify the lock holder:










Scenario
Analysis commands (SAP Note 1969700)
Details




transactional locks
SQL: “HANA_Locks_Transactional_LockWaits”

SQL: “HANA_Locks_Transactional_LockWaits_History”		
In case of transactional locks the lock holder is available via M_BLOCKED_TRANSACTIONS / HOST_BLOCKED_TRANSACTIONS (LOCK_OWNER_* columns).




(some) internal locks
SQL: “HANA_Threads_CurrentThreads”

SQL: “HANA_Threads_ThreadSamples_FilterAndAggregation”		
A subset ofinternal locks is displayed in column LOCKS_OWNED of M_SERVICE_THREADS (Rev. 1.00.120 and higher) and M_SERVICE_THREAD_SAMPLES / HOST_SERVICE_THREAD_SAMPLES (Rev. 1.00.122.08 and higher).




ConsistentChangeLock

The section [SAVEPOINT_SHAREDLOCK_OWNERS] of runtime dumps (SAP Note 2400007) provides an overview of threads holding a ConsistentChangeLock.




index handle waits

Starting with SAP HANA 1.00.120 it is possible to determine index handle lock dependencies via “hdbcons ‘indexmanager waitgraph -w'” and via runtime dump (SAP Note 2400007) section [INDEXMANAGER_WAITGRAPH]. See “Acquire Index Access” above for more information.




general
SQL: “HANA_Threads_CurrentThreads”

SQL: “HANA_Threads_ThreadSamples_FilterAndAggregation”

SQL: “HANA_Threads_Callstacks”		
If none of the above direct options exist to determine the lock holder, you have to have a close look at concurrent thread activities (SAP Note 2114710) and their call stacks (SAP Note 2313619) and watch out for usual suspects (as e.g. documented in “How can internal lock waits be analyzed?” above).
























Keywords








SAP HANA lock enqueue exclusive shared wait mutex semaphore barrier sleep conditional variable




  




  



































Product























Product or Product Version








SAP HANA, platform edition all versions





























References










This document refers to



















Number
Title








2400007
FAQ: SAP HANA Runtime Dumps






2400005
FAQ: SAP HANA Persistence






2399979
How-To: Configuring automatic SAP HANA Data Collection with SAP HANASitter






2380176
FAQ: SAP HANA Database Trace






2313619
How-To: Generating SAP HANA Call Stacks






2222250
FAQ: SAP HANA Workload Management






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






2222200
FAQ: SAP HANA Network






2215131
Alert Log Switch Wait Count Ratio






2186744
FAQ: SAP HANA Parameters






2180165
FAQ: SAP HANA Expensive Statements Trace






2180119
FAQ: SAP HANA Smart Data Access






2169283
FAQ: SAP HANA Garbage Collection






2160391
FAQ: SAP HANA Indexes






2127458
FAQ: SAP HANA Loads and Unloads






2124112
FAQ: SAP HANA Parsing






2114710
FAQ: SAP HANA Threads and Thread Samples






2112604
FAQ: SAP HANA Compression






2100040
FAQ: SAP HANA CPU






2100009
FAQ: SAP HANA Savepoints


























This document is referenced by



















Number
Title








2169283
FAQ: SAP HANA Garbage Collection






2000002
FAQ: SAP HANA SQL Optimization






2380176
FAQ: SAP HANA Database Trace






2160391
FAQ: SAP HANA Indexes






2114710
FAQ: SAP HANA Threads and Thread Samples






2399990
How-To: Analyzing ABAP Short Dumps in SAP HANA Environments






1999997
FAQ: SAP HANA Memory






1999993
How-To: Interpreting SAP HANA Mini Check Results






2399979
How-To: Configuring automatic SAP HANA Data Collection with SAP HANASitter






2180165
FAQ: SAP HANA Expensive Statements Trace






2044468
FAQ: SAP HANA Partitioning






2154870
How-To: Understanding and defining SAP HANA Limitations






2000003
FAQ: SAP HANA






2222217
How-To: Troubleshooting SAP HANA Startup Times






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






2100040
FAQ: SAP HANA CPU






2220627
FAQ: SAP HANA LOBs






2119087
How-To: Configuring SAP HANA Traces






2222277
FAQ: SAP HANA Column Store and Row Store






2100010
SAP HANA: Popular Misconceptions
























  




  














          
        


        
        

      


    

  













Tags: 






	You may also like...









	
					
		
	
		

		
Leave a Reply 
Your email address will not be published. Required fields are marked *