| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
# ⚠️ Breaking
Many deprecated queue config options are removed (actually, they should
have been removed in 1.18/1.19).
If you see the fatal message when starting Gitea: "Please update your
app.ini to remove deprecated config options", please follow the error
messages to remove these options from your app.ini.
Example:
```
2023/05/06 19:39:22 [E] Removed queue option: `[indexer].ISSUE_INDEXER_QUEUE_TYPE`. Use new options in `[queue.issue_indexer]`
2023/05/06 19:39:22 [E] Removed queue option: `[indexer].UPDATE_BUFFER_LEN`. Use new options in `[queue.issue_indexer]`
2023/05/06 19:39:22 [F] Please update your app.ini to remove deprecated config options
```
Many options in `[queue]` are are dropped, including:
`WRAP_IF_NECESSARY`, `MAX_ATTEMPTS`, `TIMEOUT`, `WORKERS`,
`BLOCK_TIMEOUT`, `BOOST_TIMEOUT`, `BOOST_WORKERS`, they can be removed
from app.ini.
# The problem
The old queue package has some legacy problems:
* complexity: I doubt few people could tell how it works.
* maintainability: Too many channels and mutex/cond are mixed together,
too many different structs/interfaces depends each other.
* stability: due to the complexity & maintainability, sometimes there
are strange bugs and difficult to debug, and some code doesn't have test
(indeed some code is difficult to test because a lot of things are mixed
together).
* general applicability: although it is called "queue", its behavior is
not a well-known queue.
* scalability: it doesn't seem easy to make it work with a cluster
without breaking its behaviors.
It came from some very old code to "avoid breaking", however, its
technical debt is too heavy now. It's a good time to introduce a better
"queue" package.
# The new queue package
It keeps using old config and concept as much as possible.
* It only contains two major kinds of concepts:
* The "base queue": channel, levelqueue, redis
* They have the same abstraction, the same interface, and they are
tested by the same testing code.
* The "WokerPoolQueue", it uses the "base queue" to provide "worker
pool" function, calls the "handler" to process the data in the base
queue.
* The new code doesn't do "PushBack"
* Think about a queue with many workers, the "PushBack" can't guarantee
the order for re-queued unhandled items, so in new code it just does
"normal push"
* The new code doesn't do "pause/resume"
* The "pause/resume" was designed to handle some handler's failure: eg:
document indexer (elasticsearch) is down
* If a queue is paused for long time, either the producers blocks or the
new items are dropped.
* The new code doesn't do such "pause/resume" trick, it's not a common
queue's behavior and it doesn't help much.
* If there are unhandled items, the "push" function just blocks for a
few seconds and then re-queue them and retry.
* The new code doesn't do "worker booster"
* Gitea's queue's handlers are light functions, the cost is only the
go-routine, so it doesn't make sense to "boost" them.
* The new code only use "max worker number" to limit the concurrent
workers.
* The new "Push" never blocks forever
* Instead of creating more and more blocking goroutines, return an error
is more friendly to the server and to the end user.
There are more details in code comments: eg: the "Flush" problem, the
strange "code.index" hanging problem, the "immediate" queue problem.
Almost ready for review.
TODO:
* [x] add some necessary comments during review
* [x] add some more tests if necessary
* [x] update documents and config options
* [x] test max worker / active worker
* [x] re-run the CI tasks to see whether any test is flaky
* [x] improve the `handleOldLengthConfiguration` to provide more
friendly messages
* [x] fine tune default config values (eg: length?)
## Code coverage:
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There are a few places in FlushQueueWithContext which make an incorrect
assumption about how `select` on multiple channels works.
The problem is best expressed by looking at the following example:
```go
package main
import "fmt"
func main() {
closedChan := make(chan struct{})
close(closedChan)
toClose := make(chan struct{})
count := 0
for {
select {
case <-closedChan:
count++
fmt.Println(count)
if count == 2 {
close(toClose)
}
case <-toClose:
return
}
}
}
```
This PR double-checks that the contexts are closed outside of checking
if there is data in the dataChan. It also rationalises the WorkerPool
FlushWithContext because the previous implementation failed to handle
pausing correctly. This will probably fix the underlying problem in
#22145
Fix #22145
Signed-off-by: Andrew Thornton <art27@cantab.net>
Signed-off-by: Andrew Thornton <art27@cantab.net>
|
| |
|
|
|
|
|
|
|
| |
Change all license headers to comply with REUSE specification.
Fix #16132
Co-authored-by: flynnnnnnnnnn <flynnnnnnnnnn@github>
Co-authored-by: John Olheiser <john.olheiser@gmail.com>
|
| |
|
|
|
|
|
|
|
| |
- Doing 64-bit atomic operations on 32-bit machines is a bit tricky by
golang, as they can only be done under certain set of
conditions(https://pkg.go.dev/sync/atomic#pkg-note-BUG).
- This PR fixes such case whereby the conditions weren't met, it moves
the int64 to the first field of the struct, which will 64-bit operations
happening on this property on 32-bit machines.
- Resolves #19518
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Continues on from #19202.
Following the addition of pprof labels we can now more easily understand the relationship between a goroutine and the requests that spawn them.
This PR takes advantage of the labels and adds a few others, then provides a mechanism for the monitoring page to query the pprof goroutine profile.
The binary profile that results from this profile is immediately piped in to the google library for parsing this and then stack traces are formed for the goroutines.
If the goroutine is within a context or has been created from a goroutine within a process context it will acquire the process description labels for that process.
The goroutines are mapped with there associate pids and any that do not have an associated pid are placed in a group at the bottom as unbound.
In this way we should be able to more easily examine goroutines that have been stuck.
A manager command `gitea manager processes` is also provided that can export the processes (with or without stacktraces) to the command line.
Signed-off-by: Andrew Thornton <art27@cantab.net>
|
| |
|
|
|
|
|
|
|
| |
Add number in queue status to the monitor page so that administrators can
assess how much work is left to be done in the queues.
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Simplify Boost/Pause logic
#18658 has added a check to see if we need to boost because there is still work to do
however the check is slightly complex and not ideal. There's no point boosting if
the queue is paused or can't scale. Therefore merge the two selects into one and add
a check to p.paused.
Signed-off-by: Andrew Thornton <art27@cantab.net>
* And on resume add a zeroboost if necessary
Signed-off-by: Andrew Thornton <art27@cantab.net>
* simplify
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Lauris BH <lauris@nix.lv>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Restart zero worker if there is still work to do
It is possible for the zero worker to timeout before all the work is finished.
This may mean that work may take a long time to complete because a worker will only
be induced on repushing.
Also ensure that requested count is reset after pulls and push mirror sync requests and add some more trace logging to the queue push.
Fix #18607
Signed-off-by: Andrew Thornton <art27@cantab.net>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(#18593)
* Only attempt to flush queue if the underlying worker pool is not finished
There is a possible race whereby a worker pool could be cancelled but yet the
underlying queue is not empty. This will lead to flush-all cycling because it
cannot empty the pool.
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Apply suggestions from code review
Co-authored-by: Gusted <williamzijl7@hotmail.com>
Co-authored-by: Gusted <williamzijl7@hotmail.com>
|
| |
|
|
|
|
|
|
|
|
| |
* Change some logging levels
* PlainTextWithBytes - 4xx/5xx this should just be TRACE
* notFoundInternal - the "error" here is too noisy and should be DEBUG
* WorkerPool - Worker pool scaling messages are normal and should be DEBUG
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Prevent deadlocks in persistable channel pause test
Because of reuse of the old paused/resumed channels in this test there
was a potential for deadlock. This PR ensures that the channels are always
reobtained.
It further adds some control code to detect hangs in future - and it
ensures that the pausing warning is not shown on shutdown.
Signed-off-by: Andrew Thornton <art27@cantab.net>
* do not warn but do pause
Signed-off-by: Andrew Thornton <art27@cantab.net>
|
| |
|
|
|
|
|
| |
Implements the Pausable interface on WrappedQueues and PersistableChannelUniqueQueues
Reference #15928
Signed-off-by: Andrew Thornton art27@cantab.net
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Start adding mechanism to return unhandled data
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Create pushback interface
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Add Pausable interface to WorkerPool and Manager
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Implement Pausable and PushBack for the bytefifos
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Implement Pausable and Pushback for ChannelQueues and ChannelUniqueQueues
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Wire in UI for pausing
Signed-off-by: Andrew Thornton <art27@cantab.net>
* add testcases and fix a few issues
Signed-off-by: Andrew Thornton <art27@cantab.net>
* fix build
Signed-off-by: Andrew Thornton <art27@cantab.net>
* prevent "race" in the test
Signed-off-by: Andrew Thornton <art27@cantab.net>
* fix jsoniter mismerge
Signed-off-by: Andrew Thornton <art27@cantab.net>
* fix conflicts
Signed-off-by: Andrew Thornton <art27@cantab.net>
* fix format
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Add warnings for no worker configurations and prevent data-loss with redis/levelqueue
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Use StopTimer
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Lauris BH <lauris@nix.lv>
Co-authored-by: 6543 <6543@obermui.de>
Co-authored-by: techknowlogick <techknowlogick@gitea.io>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
|
| |
|
|
|
|
|
|
|
|
|
| |
* gofumpt -w -l .
* gofumpt -w -l -extra .
* Add linter
* manual fix
* change make fmt
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
level queue, reduce goroutines etc (#15693)
* move shutdownfns, terminatefns and hammerfns out of separate goroutines
Coalesce the shutdownfns etc into a list of functions that get run at shutdown
rather then have them run at goroutines blocked on selects.
This may help reduce the background select/poll load in certain
configurations.
* The LevelDB queues can actually wait on empty instead of polling
Slight refactor to cause leveldb queues to wait on empty instead of polling.
* Shutdown the shadow level queue once it is empty
* Remove bytefifo additional goroutine for readToChan as it can just be run in run
* Remove additional removeWorkers goroutine for workers
* Simplify the AtShutdown and AtTerminate functions and add Channel Flusher
* Add shutdown flusher to CUQ
* move persistable channel shutdown stuff to Shutdown Fn
* Ensure that UPCQ has the correct config
* handle shutdown during the flushing
* reduce risk of race between zeroBoost and addWorkers
* prevent double shutdown
Signed-off-by: Andrew Thornton <art27@cantab.net>
|
| |
|
|
|
|
|
|
|
|
|
| |
* The leveldb shadow queue of a persistable channel queue should always start with 0
workers and just use boost to add additional workers if necessary.
* create a zero boost so that if there are no workers in a pool - boost to start the workers
* actually set timeout appropriately on boosted workers
Signed-off-by: Andrew Thornton <art27@cantab.net>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There is a potential memory leak in `Workerpool` due to the intricacies of
`time.Timer` stopping.
Whenever a `time.Timer` is `Stop`ped its channel must be cleared using a
`select` if the result of the `Stop()` is `false`.
Unfortunately in `Workerpool` these were checked the wrong way round.
However, there were a few other places that were not being checked.
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: techknowlogick <techknowlogick@gitea.io>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
|
| |
|
|
|
|
| |
* Prevent deadlock on boost
* Force a boost in testchannelqueue
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Make WorkerPools and Queues flushable
Adds Flush methods to Queues and the WorkerPool
Further abstracts the WorkerPool
Adds a final step to Flush the queues in the defer from PrintCurrentTest
Fixes an issue with Settings inheritance in queues
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Change to for loop
* Add IsEmpty and begin just making the queues composed WorkerPools
* subsume workerpool into the queues and create a flushable interface
* Add manager command
* Move flushall to queue.Manager and add to testlogger
* As per @guillep2k
* as per @guillep2k
* Just make queues all implement flushable and clean up the wrapped queue flushes
* cope with no timeout
Co-authored-by: Lauris BH <lauris@nix.lv>
|
|
|
* Queue: Add generic graceful queues with settings
* Queue & Setting: Add worker pool implementation
* Queue: Add worker settings
* Queue: Make resizing worker pools
* Queue: Add name variable to queues
* Queue: Add monitoring
* Queue: Improve logging
* Issues: Gracefulise the issues indexer
Remove the old now unused specific queues
* Task: Move to generic queue and gracefulise
* Issues: Standardise the issues indexer queue settings
* Fix test
* Queue: Allow Redis to connect to unix
* Prevent deadlock during early shutdown of issue indexer
* Add MaxWorker settings to queues
* Merge branch 'master' into graceful-queues
* Update modules/indexer/issues/indexer.go
Co-Authored-By: guillep2k <18600385+guillep2k@users.noreply.github.com>
* Update modules/indexer/issues/indexer.go
Co-Authored-By: guillep2k <18600385+guillep2k@users.noreply.github.com>
* Update modules/queue/queue_channel.go
Co-Authored-By: guillep2k <18600385+guillep2k@users.noreply.github.com>
* Update modules/queue/queue_disk.go
* Update modules/queue/queue_disk_channel.go
Co-Authored-By: guillep2k <18600385+guillep2k@users.noreply.github.com>
* Rename queue.Description to queue.ManagedQueue as per @guillep2k
* Cancel pool workers when removed
* Remove dependency on queue from setting
* Update modules/queue/queue_redis.go
Co-Authored-By: guillep2k <18600385+guillep2k@users.noreply.github.com>
* As per @guillep2k add mutex locks on shutdown/terminate
* move unlocking out of setInternal
* Add warning if number of workers < 0
* Small changes as per @guillep2k
* No redis host specified not found
* Clean up documentation for queues
* Update docs/content/doc/advanced/config-cheat-sheet.en-us.md
* Update modules/indexer/issues/indexer_test.go
* Ensure that persistable channel queue is added to manager
* Rename QUEUE_NAME REDIS_QUEUE_NAME
* Revert "Rename QUEUE_NAME REDIS_QUEUE_NAME"
This reverts commit 1f83b4fc9b9dabda186257b38c265fe7012f90df.
Co-authored-by: guillep2k <18600385+guillep2k@users.noreply.github.com>
Co-authored-by: Lauris BH <lauris@nix.lv>
Co-authored-by: techknowlogick <matti@mdranta.net>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
|
