Modify the implementation of MemTracker:
1. Simplify a lot of useless logic;
2. Added MemTrackerTaskPool, as the ancestor of all query and import trackers, This is used to track the local memory usage of all tasks executing;
3. Add cosume/release cache, trigger a cosume/release when the memory accumulation exceeds the parameter mem_tracker_consume_min_size_bytes;
4. Add a new memory leak detection mode (Experimental feature), throw an exception when the remaining statistical value is greater than the specified range when the MemTracker is destructed, and print the accurate statistical value in HTTP, the parameter memory_leak_detection
5. Added Virtual MemTracker, cosume/release will not sync to parent. It will be used when introducing TCMalloc Hook to record memory later, to record the specified memory independently;
6. Modify the GC logic, register the buffer cached in DiskIoMgr as a GC function, and add other GC functions later;
7. Change the global root node from Root MemTracker to Process MemTracker, and remove Process MemTracker in exec_env;
8. Modify the macro that detects whether the memory has reached the upper limit, modify the parameters and default behavior of creating MemTracker, modify the error message format in mem_limit_exceeded, extend and apply transfer_to, remove Metric in MemTracker, etc.;
Modify where MemTracker is used:
1. MemPool adds a constructor to create a temporary tracker to avoid a lot of redundant code;
2. Added trackers for global objects such as ChunkAllocator and StorageEngine;
3. Added more fine-grained trackers such as ExprContext;
4. RuntimeState removes FragmentMemTracker, that is, PlanFragmentExecutor mem_tracker, which was previously used for independent statistical scan process memory, and replaces it with _scanner_mem_tracker in OlapScanNode;
5. MemTracker is no longer recorded in ReservationTracker, and ReservationTracker will be removed later;
As described in #8120, a large number of rowset meta remain in rocksdb, which may be generated by:
1. drop tablet
The drop tablet task itself just sets the state of the tablet meta to `SHUTDOWN`
and moves the tablet to `_shutdown_tablets` vector then the background thread
will periodically clean up the tablet in `_shutdown_tablets` (that's why even if we execute
the `drop table xx force`, the tablet may be delayed by 10min to 1 hour before it goes into the trash directory).
The regular cleanup thread in the background saves the complete tablet meta as a `.hdr` file
when deleting the tablet, and then moves it to the trash directory along with the data files.
But this process does not process the rowset meta (before doing the checkpoint of the tablet meta,
the rowset meta is stored independently in rocksdb as a key-value). So this results in a residual rowset meta.
2. clone task
The clone task may migrate back and forth between BEs, which may result in a situation
where the tablet id is the same on the BE, but the tablet uuid is different.
This leads to some rowset meta can not find the corresponding tablet, but there is no thread
to process these rowsets, and eventually lead to residual.
This is PR, I handled it in the regular cleanup thread with method `_clean_unused_rowset_metas()`.
I did not delete rowset meta along with "drop tablet" task, because "drop tablet" itself is not a synchronous operation.
It also relies on a background thread to clean up the tablet periodically.
So I put this operation in the background cleanup thread.
For the first, we need to make a parameter to discribe the data is local or remote.
At then, we need to support some basic function to support the operation for remote storage.
1. replace all boost::shared_ptr to std::shared_ptr
2. replace all boost::scopted_ptr to std::unique_ptr
3. replace all boost::scoped_array to std::unique<T[]>
4. replace all boost:thread to std::thread
1. `StorageEngine::_delete_tablets_on_unused_root_path` will try to obtain tablet shard write lock in `TabletManager`
```
StorageEngine::_delete_tablets_on_unused_root_path
TabletManager::drop_tablets_on_error_root_path
obtain each tablet shard's write lock
```
2. `TabletManager::build_all_report_tablets_info` and other methods will obtain tablet shard read lock frequently.
So, `StorageEngine::_delete_tablets_on_unused_root_path` will hold `_store_lock` for a long time.
This will make it difficult for other threads to get write `_store_lock`, such as `StorageEngine::get_stores_for_create_tablet`
`drop_tablets_on_error_root_path` is a small probability event, `TabletManager::drop_tablets_on_error_root_path` should return when its param `tablet_info_vec` is empty
random_shuflle will generate same random sequence when call multiple times,
although we use twice random, but when there is no change in the size relationship
between the adjacent numbers, the result of the second shuffle will not change either
To avoid showing too many memtracker on BE web pages.
The MemTracker level now has 3 levels: OVERVIEW, TASK and VERBOSE.
OVERVIEW Mainly used for main memory consumption module such as Query/Load/Metadata.
TASK is mainly used to record the memory overhead of a single task such as a single query, load, and compaction task.
VERBOSE is used for other more detailed memtrackers.
We created multiple rowset readers to read data of one tablet,
after one rowset reader has reached EOF, it can be released to
reduce resource (typically memory) consumption.
As the same, we can release segment reader when it reach EOF.
The version information of the tablet will be stored in the memory
in an adjacency graph data structure.
And as the new version is written and the old version is deleted,
the data structure will begin to have empty vertex with no edge associations(orphan vertex).
These orphan vertexs should be removed somehow.
1. Add /api/compaction/run_status to show the running compaction tasks.
2. Support do base and cumulative compaction for one tablet at same time.
3. Modify some log level.
4. Add a feedback document.
Currently Tablet meta checkpoint is a memory-exhausted operation.
If a host has 12 disks, it will start 12 threads to do tablet meta checkpoint.
In our experience, the data size of one tablet can be as high as 2G.
If 12 threads do the checkpoint at the same time, it maybe cause OOM.
Therefore, this PR try to solve this problem.
Firstly, it only start one thread to produce table meta checkpoint tasks.
Secondly, it creates a thread pool to handle these tasks.
You can configure the size of the thread pool to control the parallelism in case of OOM.
It is a producer-customer model.
Record finished stream load job (both successful job and failed job) into audit log
so that we can see when the stream load job was executed and check the details of stream load jobs.
MemTracker can provide memory consumption for us to find out which
module consume more memory, but it's just a current value, this patch
add metrics for some large memory consumers, then we can find out
which module consume more memory in timeline, it would be useful to
troubleshoot OOM problems and optimize configs.
* [Enhance] Make MemTracker more accurate (#5515)
This PR main about:
1. Improve the readability of MemTrackers' name
2. Add the MemTracker of:
* Load
* Compaction
* SchemaChange
* StoragePageCache
* TabletManager
3. Change SchemaChange to a Singleon
* revise some code for Code Review
* change the name of mem_tracker
* keep reader_context have the same lifetime of rowset_reader in schema change.
* change vlog notice to log(warning) in schema change
* update gcc to gcc 10 and support c++17
update brpc to 0.9.7
update boost to 1.73
remove third-party boost 1.54 for mysql
* update cmake version
* ignore jdk version
* remove unused patch
* avoid use SYS_getrandom call
* [Enhance] Sort directories by available space when do trash sweep
In the case when one disk is about to be full, we want to sweep trash
data on this disk as quickly as possible. The currently trash sweep
function is to remove trashed files order by path's name, however, disk
data directories may have some large different available space because
of the load balance algorithm, this patch improve it to remove files by
directories' available space.
* add log
At present, the application of vlog in the code is quite confusing.
It is inherited from impala VLOG_XX format, and there is also VLOG(number) format.
VLOG(number) format does not have a unified specification, so this pr standardizes the use of VLOG
ThreadSanitizer, aka TSAN, is a useful tool to detect multi-thread
problems, such as data race, mutex problems, etc.
We should detect TSAN problems for Doris BE, both unit tests and
server should pass through TSAN mode, to make Doris more robustness.
This is the very beginning patch to fix TSAN problems, and some
difficult problems are suppressed in file 'tsan_suppressions', you
can suppress these problems by setting:
export TSAN_OPTIONS="suppressions=tsan_suppressions"
before running:
`BUILD_TYPE=tsan ./run-be-ut.sh --run`
Add trace for create tablet tasks, it's a useful tool for admin to find
out the bottleneck when create tablets timeouted.
For example, admin could enlarge 'tablet_map_shard_size' when found
'got tablets shard lock' procedure cost too much time.
The current compaction mechanism is that there is a producer thread that has been producing compaction tasks,
and the selected tablet must apply for `permits`.
When a tablet could hold `permits`, compaction task for this tablet will be submitted to thread pool.
We take compaction score as `permits` which is used for limiting memory consumption.
However, `pick_rowset_to_compaction()` will be executed before the file merge in compaction thread,
and the number of segment files that actually perform the merge operation is smaller than compaction score.
In addition, it is also possible that compaction task exits directly because the tablet doesn't meet
the requirements of compaction.
This patch optimizes and refactors the code of compaction, so that we can execute 'pick rowsets'
before applying for permits for a compaction task, calculate the number of segment files that actually
participate in the merge operation, and take this number as `permits`.
We create some objects which are only used for unit tests, it's not necessary,
and it may cause create duplicate instances for some classes.
This patch remove unnecessary instance of class BlockManager and StoragePageCache.
Currently, there are M threads to do base compaction and N threads to do cumulative compaction for each disk.
Too many compaction tasks may run out of memory, so the max concurrency of running compaction tasks
is limited by semaphore.
If the running threads cost too much memory, we can't defense it. In addition, reducing concurrency to avoid OOM
will lead to some compaction tasks can't be executed in time and we may encounter more heavy compaction.
Therefore, concurrency limitation is not enough.
The strategy proposed in #3624 may be effective to solve the OOM.
A CompactionPermitLimiter is used for compaction limitation, and use single-producer/multi-consumer model.
Producer will try to generate compaction tasks and acquire `permits` for each task.
The compaction task which can hold `permits` will be executed in thread pool and each finished task will
release its `permits`.
`permits` should be applied for before a compaction task can execute. When the sum of `permits`
held by executing compaction tasks reaches a threshold, subsequent compaction task will be no longer allowed,
until some `permits` are released. Tablet compaction score is used as `permits` of compaction task here.
To some extent, memory consumption can be limited by setting appropriate `permits` threshold.
The tablet and disk information reporting threads need to report to the FE periodically.
At the same time these two reporting threads will also be triggered by certain events.
The modification in PR #4440 caused these two threads to be triggered only by events,
and could not report regularly.
BE can not graceful exit because some threads are running in endless
loop. This patch do the following optimization:
- Use the well encapsulated Thread and ThreadPool instead of std::thread
and std::vector<std::thread>
- Use CountDownLatch in thread's loop condition to avoid endless loop
- Introduce a new class Daemon for daemon works, like tcmalloc_gc,
memory_maintenance and calculate_metrics
- Decouple statistics type TaskWorkerPool and StorageEngine notification
by submit tasks to TaskWorkerPool's queue
- Reorder objects' stop and deconstruct in main(), i.e. stop network
services at first, then internal services
- Use libevent in pthreads mode, by calling evthread_use_pthreads(),
then EvHttpServer can exit gracefully in multi-threads
- Call brpc::Server's Stop() and ClearServices() explicitly
Sometimes we want to detect the hotspot of a cluster, for example, hot scanned tablet, hot wrote tablet,
but we have no insight about tablets in the cluster.
This patch introduce tablet level metrics to help to achieve this object, now support 4 metrics on tablets: `query_scan_bytes `, `query_scan_rows `, `flush_bytes `, `flush_count `.
However, one BE may holds hundreds of thousands of tablets, so I add a parameter for the metrics HTTP request,
and not return tablet level metrics by default.
