vanus-labs / vanus Goto Github PK

vsctl subscription info add response param, such as offset

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

2. What did you expect to see? (Required)

The number of eventlog need equal to the number of eventbug.

3. What did you see instead (Required)

Inconsistent number of eventbus and eventlog。

4. What is your Vanus version? (Required)

Enhancement

add minikube profile for vanus

trigger use client sdk read event has rpc error

time="2022-08-11T08:17:00.798155619Z" level=warning msg="read event error" error="rpc error: code = Canceled desc = context canceled" eventlog_id="vanus:///eventlog/1660202185087308970?eventbus=performance-1&controllers=172.31.54.156:2048" offset=124611

Bug Report

store startup failed.

`
[centos@ip-172-31-59-20 vanus]$ kubectl get pod -n vanus
NAME READY STATUS RESTARTS AGE
vanus-controller-0 1/1 Running 0 7m22s
vanus-controller-1 1/1 Running 0 7m16s
vanus-controller-2 1/1 Running 0 7m11s
vanus-gateway-78d757d594-xtt69 1/1 Running 0 7m23s
vanus-store-0 0/1 CrashLoopBackOff 1 (4s ago) 6s
vanus-trigger-6996cfbdcb-c2zg7 1/1 Running 3 (6m55s ago) 7m23s
[centos@ip-172-31-59-20 vanus]$ kubectl logs -n vanus vanus-store-0
time="2022-08-11T07:28:08.676568019Z" level=debug msg="logger level has been set" log_level=DEBUG
time="2022-08-11T07:28:08.678628336Z" level=error msg="Create iouring failed." error="iouring_setup: function not implemented"
panic: iouring_setup: function not implemented

goroutine 1 [running]:
github.com/linkall-labs/vanus/internal/store/io.NewURing()
github.com/linkall-labs/vanus/internal/store/io/uring.go:54 +0x1dd
github.com/linkall-labs/vanus/internal/store/wal.defaultIOEngine(...)
github.com/linkall-labs/vanus/internal/store/wal/config_linux.go:23
github.com/linkall-labs/vanus/internal/store/wal.defaultWALConfig()
github.com/linkall-labs/vanus/internal/store/wal/config.go:57 +0x45
github.com/linkall-labs/vanus/internal/store/wal.makeConfig({0xc0000685e0, 0x3, 0x44a134})
github.com/linkall-labs/vanus/internal/store/wal/config.go:65 +0x89
github.com/linkall-labs/vanus/internal/store/wal.Open({0xc000029690, 0xa}, {0xc0000685e0, 0xc00000e3f8, 0x0})
github.com/linkall-labs/vanus/internal/store/wal/wal.go:104 +0x4c
github.com/linkall-labs/vanus/internal/store/meta.RecoverSyncStore({{0x0, {{0xc000029520, 0x5}}}}, {0xc000029690, 0xa})
github.com/linkall-labs/vanus/internal/store/meta/sync.go:178 +0x233
github.com/linkall-labs/vanus/internal/store/segment.(*server).recover(0xc00000a3c0, {0xa7f110, 0xc000028048})
github.com/linkall-labs/vanus/internal/store/segment/recovery.go:32 +0xd3
github.com/linkall-labs/vanus/internal/store/segment.(*server).Initialize(0xc00000a3c0, {0xa7f110, 0xc000028048})
github.com/linkall-labs/vanus/internal/store/segment/server.go:192 +0x2c
main.main()
./main.go:56 +0x3f0
[centos@ip-172-31-59-20 vanus]$
`
Should be caused by the kernel version.

1. Minimal reproduce step (Required)

2. What did you expect to see? (Required)

3. What did you see instead (Required)

4. What is your Vanus version? (Required)

v0.2.0

Enhancement

the logic of save offset to etcd with sync will slow down performance

Enhancement

due to unstable ut, which causes CI action to fail randomly, see https://github.com/linkall-labs/vanus/pull/125/checks?check_run_id=7068742869 for detail

Enhancement

• Delete compacted data of WAL.
• Unit tests for compaction.

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

sink lost some event messages in e2e test.
I sent 10000 event messages, but sink received only 4930.
https://github.com/linkall-labs/vanus/runs/7212590689?check_suite_focus=true

Run kubectl get pod -n vanus | grep quick-display | awk '{print $1}' | xargs kubectl logs -n vanus | grep -n "total" | wc -l
4930
failed

2. What did you expect to see? (Required)

3. What did you see instead (Required)

4. What is your Vanus version? (Required)

Enhancement

add recover logic to prevent other package panic causing the timer to restart.
`
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x10 pc=0xad9cba]

goroutine 8144 [running]:
github.com/linkall-labs/vanus/client/internal/vanus/eventlog.(*logSegment).EndOffset(...)
/home/centos/linkall/vanus/client/internal/vanus/eventlog/log_segment.go:98
github.com/linkall-labs/vanus/client/internal/vanus/eventlog.(*logReader).Read(0xc00081c240, {0xdd6210, 0xc000c87a00}, 0x56e0)
/home/centos/linkall/vanus/client/internal/vanus/eventlog/distributed_eventlog.go:392 +0x15a
github.com/linkall-labs/vanus/internal/timer/timingwheel.(*bucket).getEvent(0xc000423c00, {0xdd6210, 0xc000c87a00}, 0x2e32)
/home/centos/linkall/vanus/internal/timer/timingwheel/bucket.go:394 +0x1c6
github.com/linkall-labs/vanus/internal/timer/timingwheel.(*bucket).fetchEventFromOverflowWheelAdvance(0xc000423c00, {0xdd6210, 0xc000c87a00}, 0xc0003b7700)
/home/centos/linkall/vanus/internal/timer/timingwheel/bucket.go:295 +0x96
github.com/linkall-labs/vanus/internal/timer/timingwheel.(*timingWheelElement).fetchEventFromOverflowWheelAdvance.func1()
/home/centos/linkall/vanus/internal/timer/timingwheel/timingwheel.go:617 +0x190
created by github.com/linkall-labs/vanus/internal/timer/timingwheel.(*timingWheelElement).fetchEventFromOverflowWheelAdvance
/home/centos/linkall/vanus/internal/timer/timingwheel/timingwheel.go:602 +0x257
`

Enhancement

Add manual that how to configure grafana+prometheus in local

Bug Report

vanus-controller container restarts continuously

ubuntu@ip-172-31-57-192:~/jiangkai/projects/file$ kubectl get pod -n vanus
NAME                             READY   STATUS    RESTARTS         AGE
vanus-controller-0               1/1     Running   4 (8m40s ago)    48m
vanus-controller-1               1/1     Running   4 (8m38s ago)    48m
vanus-controller-2               1/1     Running   4 (8m28s ago)    48m

vanus-controller-0.log

1. Minimal reproduce step (Required)

deploy all-in-one.yaml

2. What did you expect to see? (Required)

container not restarted

3. What did you see instead (Required)

container restarts continuously

4. What is your Vanus version? (Required)

image tag: cae0d9b

Enhancement

code optimization for access field of list directly

Enhancement

Currently, the E2E test uses display as a sink to verify the consumption of events, but the way of querying logs is not flexible enough, so an enhanced sink needs to be implemented.

This sink can reflect the event consumption in a more flexible way.

Enhancement

add release specification

Bug Report

timing message consume failed when eventbus not exist.
time="2022-08-11T08:32:20.010086035Z" level=error msg="lookup readable logs failed" error="rpc error: code = Unknown desc = {\"code\":\"RESOURCE_NOT_FOUND\",\"message\":\"eventbus not found\"}"

1. Minimal reproduce step (Required)

2. What did you expect to see? (Required)

3. What did you see instead (Required)

4. What is your Vanus version? (Required)

Enhancement

Adding Panels of metrics that were added in v0.2

now reset offset step is below:

1. stop trigger
2. get offset from eventbus client
3. commit offset to trigger controller
4. start trigger

if fail on step 2 or 3 or 4, the trigger is stopped and does not start again, have another solution no need to stop the trigger, or have a daemon thread monitor trigger that needs running but does not start will start the trigger

Enhancement

to Delu's advice in #138

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

2. What did you expect to see? (Required)

3. What did you see instead (Required)

subscriptionScheduler stop has panic

4. What is your Vanus version? (Required)

0.1.0

now vsctl subscription create from has 3 params fromLatest, fromEarliest, fromTime, optimize use a param from, for example:

--from latest
--from earliest
--from 2022-01-01T01:01:01Z

Enhancement

support dind in quickstart

Bug Report

vsctl event put returns 200 but message lost

1. Minimal reproduce step (Required)

step1: vsctl event put __Timer_0_0 xxx
step2: vsctl event get __Timer_0_0 --offset 0 --number 1
step2 return 500, message not found

2. What did you expect to see? (Required)

3. What did you see instead (Required)

4. What is your Vanus version? (Required)

v0.1.2

Enhancement

use consumption pool to solve the problem of concurrent consumption performance

Enhancement

optimize fault recovery judgment logic.
The current fault recovery judgment logic uses logic of bucket start, which will lead to some error log printing, so needs to be re implemented.

Enhancement

Integrating Grafana with k8s, provide an out-of-box experience

Enhancement

supports graceful shutdown for timer
refer: https://github.com/linkall-labs/vanus/blob/main/cmd/controller/main.go#L136

Enhancement

add display to vanus-bench receive

Enhancement

reduce the latency of the schedule message is consumed when its time arrived, now is 3s.

Bug Report

vsctl event put failed and return 400

gateway log:

time="2022-08-02T15:51:25.770926378Z" level=warning msg="append to failed" error="not writable" vrn="vanus://vanus-controller-0.vanus-controller:2048/eventbus/__Timer_0_0?controllers=vanus-controller-0.vanus-controller:2048,vanus-controller-1.vanus-controller:2048,vanus-controller-2.vanus-controller:2048"

store log:

time="2022-08-02T15:51:20.867241073Z" level=warning msg="Append failed." blockID=1659441842434422563 error="not leader"
time="2022-08-02T15:51:20.873542311Z" level=warning msg="Append failed." blockID=1659441842434422563 error="not leader"
time="2022-08-02T15:51:20.875681891Z" level=warning msg="Append failed." blockID=1659441842434422563 error="not leader"
time="2022-08-02T15:51:21.736343391Z" level=debug msg=heartbeat leader="vanus-controller-1.vanus-controller.vanus.svc:2048"
time="2022-08-02T15:51:22.465173574Z" level=debug msg="Append entries to raft log." appended_index=18 block_id=1659450063760863077 entries_num=1
time="2022-08-02T15:51:22.475198983Z" level=debug msg="Persist raft hard state." block_id=1659450063760863077 hard_state="{2 1659450064068188923 18}"
time="2022-08-02T15:51:22.476955757Z" level=debug msg="Store applied offset." applied_offset=18 block_id=1659450063760863077

1. Minimal reproduce step (Required)

vsctl event put

2. What did you expect to see? (Required)

event put success and return 200

3. What did you see instead (Required)

event put failed and return 400

4. What is your Vanus version? (Required)

v0.1.2

Enhancement

In this current implementation, The SegmentServer/Trigger will shut down immediately when they start, but the Controller need some time to initialize, this will cause SegmentServer/Trigger CrashLoopBackOff when cluster init. we can sleep sometimes when starting instead of a shutdown immediately.

eventbus info support JSON output

testtest

Enhancement

TriggerWorker,SementServer,Gateway all need connect to controller leader, but the code is maintained separately ,exist code duplication，can extract a lib or func？

Enhancement

add --limit to vsctl eventbus info

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

2. What did you expect to see? (Required)

3. What did you see instead (Required)

Can't start store with corrupted data.

raft2022/06/27 06:48:12 INFO: 16f9479304b9dfed switched to configuration voters=(1655433034800422893 1655433044971246865 1655433062803036836)
raft2022/06/27 06:48:12 16f9479304b9dfed state.commit 5 is out of range [4, 4]
panic: 16f9479304b9dfed state.commit 5 is out of range [4, 4]

goroutine 1 [running]:
log.(*Logger).Panicf(0xc0001c2ee0, {0x9bdc2d, 0x40a905}, {0xc00040ab40, 0x8e1dc0, 0x0})
	log/log.go:231 +0x74
github.com/linkall-labs/vanus/raft.(*DefaultLogger).Panicf(0x16f9479304b9dfed, {0x9bdc2d, 0x0}, {0xc00040ab40, 0x0, 0x0})
	github.com/linkall-labs/vanus/[email protected]/logger.go:137 +0x26
github.com/linkall-labs/vanus/raft.(*raft).loadState(0xc00024b600, {0x3, 0x3, 0x0})
	github.com/linkall-labs/vanus/[email protected]/raft.go:1724 +0x19a
github.com/linkall-labs/vanus/raft.newRaft(0xc0001bdab8)
	github.com/linkall-labs/vanus/[email protected]/raft.go:357 +0x6de
github.com/linkall-labs/vanus/raft.NewRawNode(0xc000100870)
	github.com/linkall-labs/vanus/[email protected]/rawnode.go:49 +0x1d
github.com/linkall-labs/vanus/raft.RestartNode(0xc0001c2ee0)
	github.com/linkall-labs/vanus/[email protected]/node.go:245 +0x2b
github.com/linkall-labs/vanus/internal/store/block/replica.New({0xa5f170, 0xc000116008}, 0x16f9479304b9dfed, {0xa5f288, 0xc0002255e0}, 0xc0001c2ee0, {0x7f1838cab7b8, 0xc000200280}, 0xc000416310)
	github.com/linkall-labs/vanus/internal/store/block/replica/replica.go:131 +0x388
github.com/linkall-labs/vanus/internal/store/segment.(*server).makeReplicaWithRaftLog(0xc0002024e0, {0xa5f170, 0xc000116008}, 0x991780, {0xa5f288, 0xc0002255e0}, 0x885cc0)
	github.com/linkall-labs/vanus/internal/store/segment/server.go:692 +0xf5
github.com/linkall-labs/vanus/internal/store/segment.(*server).recoverBlocks(0xc0002024e0, {0xa5f170, 0xc000116000}, 0xc00038faa0)
	github.com/linkall-labs/vanus/internal/store/segment/recovery.go:86 +0x30e
github.com/linkall-labs/vanus/internal/store/segment.(*server).recover(0xc0002024e0, {0xa5f170, 0xc000116000})
	github.com/linkall-labs/vanus/internal/store/segment/recovery.go:50 +0x20e
github.com/linkall-labs/vanus/internal/store/segment.(*server).Initialize(0xc0002024e0, {0xa5f170, 0xc000116000})
	github.com/linkall-labs/vanus/internal/store/segment/server.go:132 +0x2c
main.main()
	./main.go:82 +0x5d3

4. What is your Vanus version? (Required)

0.1.0

Bug Report

Load timing message form overflowwheel has error log.

time="2022-08-05T02:49:37.780216419Z" level=debug msg="load event to next layer timingwheel" expiration="2022-08-05 02:50:10.555555 +0000 UTC" sourceEventbus=__Timer_2_5 time="2022-08-05T02:49:37.7840398Z" level=debug msg="end loading from overflowwheel" layer=2 pointer=3 time="2022-08-05T02:49:37.784145731Z" level=error msg="append event to failed" error="rpc error: code = Canceled desc = context canceled" eventbus=__Timer_2_6 expiration="2022-08-05 02:50:10.555555 +0000 UTC" time="2022-08-05T02:49:37.784180372Z" level=error msg="add event to eventbus failed" eventbus=__Timer_2_6 expiration="2022-08-05 02:50:10.555555 +0000 UTC" time="2022-08-05T02:49:37.78419176Z" level=debug msg="context canceled at bucket load" time="2022-08-05T02:49:38.167591239Z" level=debug msg="get event success" eventbus=__Timer_2_5 number=1 offset=0 time="2022-08-05T02:49:38.167662475Z" level=debug msg="load event to next layer timingwheel" expiration="2022-08-05 02:50:10.555555 +0000 UTC" sourceEventbus=__Timer_2_5 time="2022-08-05T02:49:38.741130645Z" level=debug msg="timingwheel pointer timer" layer=1 pointer=28 time="2022-08-05T02:49:38.810197148Z" level=debug msg="put event success" eventTime="2022-08-05 02:50:10.555555 +0000 UTC" eventbus=__Timer_1_27 time="2022-08-05T02:49:38.810235563Z" level=info msg="add event to eventbus success" eventbus=__Timer_1_27 expiration="2022-08-05 02:50:10.555555 +0000 UTC" time="2022-08-05T02:49:38.81024663Z" level=info msg="reinsert timing message success" expiration="2022-08-05 02:50:10.555555 +0000 UTC" sourceEventbus=__Timer_2_5

1. Minimal reproduce step (Required)

2. What did you expect to see? (Required)

3. What did you see instead (Required)

4. What is your Vanus version? (Required)

Enhancement

like

Bug Report

writing the message immediately after creating eventbus fails with an error "not writable"

from the log, it took about 1s to write successfully.

1. Minimal reproduce step (Required)

write messages through eventlogwriter after creating eventbus

2. What did you expect to see? (Required)

the message can be written immediately after the eventbus is created successfully

3. What did you see instead (Required)

4. What is your Vanus version? (Required)

trigger read event log too many , need reduce

time="2022-08-11T08:17:00.798155619Z" level=warning msg="read event error" error="rpc error: code = Canceled desc = context canceled" eventlog_id="vanus:///eventlog/1660202185087308970?eventbus=performance-1&controllers=172.31.54.156:2048" offset=124611

• vsctl subscription get and list table display style format
• vsctl subscription list response default does not contain the offsets
• vsctl subscription list response add limit
• vsctl subscription create param optimize

Enhancement

add version sub-command to vsctl

now config use fix ip 192.168.49.2(minikube default use) as gateway endpoint，if ip is not this, vsctl can not running，need

Enhancement

subscription support update

• update filter
• update transformation
• update sink

Enhancement

The highest layer timingwheel needs to clean up expired buckets.
Because the timingwheel at the highest layer is an data structure of limitless length array, it is necessary to clean up expired buckets after loading to release useless memory overhead.
The cleaning cycle is triggered once after each loading of the highest level timingwheel.

Enhancement

According to discussion of #117, We should reduce service in the gateway module. This issue tracks the implementation of it.

Enhancement

change --body to --data of vsctl event put to keep consistent with cloudevent