2017-02-09 15:29:17 +00:00
|
|
|
package test
|
|
|
|
|
|
Fix: maptracker race issues
This commit attempts to fix race issues in the maptracker since the
introduction of the OperationTracker.
There were two main problems:
* Duplicity tracking the state both in the state map and the opTracker
* Non atomiciy of operations with different threads being able to affect
other threads operations.
A test performing random Track/Untracks on the same Cid quickly showed
that items would sometimes stay as pin_queued or pin_unqueued. That happened
because operations could be cancelled under the hood by a different request,
while leaving the map status untouched.
It was not simply to deal with this issues without a refactoring.
First, the state map has been removed, and the operation tracker now provides
status information for any Cid. This implies that the tracker keeps all
operations and operations have a `PhaseDone`. There's also a
new `OperationRemote` type.
Secondly, operations are only created in the tracker and can only be removed
by their creators (they can be overwritten by other operations though).
Operations cannot be accessed directly and modifications are limited to setting
Error for PhaseDone operations.
After created, *Operations are queued in the pinWorker queues which handle any
status updates. This means, that, even when an operation has been removed from
the tracker, status updates will not interfere with any other newer operations.
In the maptracker, only the Unpin worker Cleans operations once processed. A
sucessful unpin is the only way that a delete() happens in the tracker map.
Otherwise, operations stay there until a newer operation for the Cid arrives
and 1) cancels the existing one 2) takes its place. The tracker refuses to
create a new operation if a similar "ongoing" operation of the same type
exists.
The final change is that Recover and RecoverAll() are not async and play by the
same rules as Track() and Untrack(), queueing the items to be recovered.
Note: for stateless pintracker, the tracker will need to Clean() operation
of type OperationPin as well, and complement the Status reported
by the tracker with those coming from IPFS.
License: MIT
Signed-off-by: Hector Sanjuan <code@hector.link>
2018-05-25 16:32:10 +00:00
|
|
|
import (
|
|
|
|
|
cid "github.com/ipfs/go-cid"
|
|
|
|
|
peer "github.com/libp2p/go-libp2p-peer"
|
|
|
|
|
)
|
2017-02-09 15:29:17 +00:00
|
|
|
|
2017-12-06 14:00:01 +00:00
|
|
|
// Common variables used all around tests.
|
2017-02-09 15:29:17 +00:00
|
|
|
var (
|
2018-04-18 04:53:41 +00:00
|
|
|
TestCid1 = "QmP63DkAFEnDYNjDYBpyNDfttu1fvUw99x1brscPzpqmmq"
|
|
|
|
|
TestCid2 = "QmP63DkAFEnDYNjDYBpyNDfttu1fvUw99x1brscPzpqmma"
|
|
|
|
|
TestCid3 = "QmP63DkAFEnDYNjDYBpyNDfttu1fvUw99x1brscPzpqmmb"
|
2018-02-15 04:59:57 +00:00
|
|
|
TestCid4 = "zb2rhiKhUepkTMw7oFfBUnChAN7ABAvg2hXUwmTBtZ6yxuc57"
|
2018-02-14 04:44:43 +00:00
|
|
|
TestCid4Data = "Cid4Data" // Cid resulting from block put NOT ipfs add
|
2018-04-18 04:53:41 +00:00
|
|
|
TestSlowCid1 = "QmP63DkAFEnDYNjDYBpyNDfttu1fvUw99x1brscPzpqmmd"
|
2017-02-15 14:16:34 +00:00
|
|
|
// ErrorCid is meant to be used as a Cid which causes errors. i.e. the
|
|
|
|
|
// ipfs mock fails when pinning this CID.
|
2018-07-23 13:56:46 +00:00
|
|
|
ErrorCid = "QmP63DkAFEnDYNjDYBpyNDfttu1fvUw99x1brscPzpqmmc"
|
2017-02-09 15:29:17 +00:00
|
|
|
TestPeerID1, _ = peer.IDB58Decode("QmXZrtE5jQwXNqCJMfHUTQkvhQ4ZAnqMnmzFMJfLewuabc")
|
|
|
|
|
TestPeerID2, _ = peer.IDB58Decode("QmUZ13osndQ5uL4tPWHXe3iBgBgq9gfewcBMSCAuMBsDJ6")
|
|
|
|
|
TestPeerID3, _ = peer.IDB58Decode("QmPGDFvBkgWhvzEK9qaTWrWurSwqXNmhnK3hgELPdZZNPa")
|
2018-01-18 02:49:35 +00:00
|
|
|
TestPeerID4, _ = peer.IDB58Decode("QmZ8naDy5mEz4GLuQwjWt9MPYqHTBbsm8tQBrNSjiq6zBc")
|
|
|
|
|
TestPeerID5, _ = peer.IDB58Decode("QmZVAo3wd8s5eTTy2kPYs34J9PvfxpKPuYsePPYGjgRRjg")
|
|
|
|
|
TestPeerID6, _ = peer.IDB58Decode("QmR8Vu6kZk7JvAN2rWVWgiduHatgBq2bb15Yyq8RRhYSbx")
|
2018-09-12 13:56:10 +00:00
|
|
|
|
|
|
|
|
TestPeerName1 = "TestPeer1"
|
|
|
|
|
TestPeerName2 = "TestPeer2"
|
|
|
|
|
TestPeerName3 = "TestPeer3"
|
|
|
|
|
TestPeerName4 = "TestPeer4"
|
|
|
|
|
TestPeerName5 = "TestPeer5"
|
|
|
|
|
TestPeerName6 = "TestPeer6"
|
2017-02-09 15:29:17 +00:00
|
|
|
)
|
Fix: maptracker race issues
This commit attempts to fix race issues in the maptracker since the
introduction of the OperationTracker.
There were two main problems:
* Duplicity tracking the state both in the state map and the opTracker
* Non atomiciy of operations with different threads being able to affect
other threads operations.
A test performing random Track/Untracks on the same Cid quickly showed
that items would sometimes stay as pin_queued or pin_unqueued. That happened
because operations could be cancelled under the hood by a different request,
while leaving the map status untouched.
It was not simply to deal with this issues without a refactoring.
First, the state map has been removed, and the operation tracker now provides
status information for any Cid. This implies that the tracker keeps all
operations and operations have a `PhaseDone`. There's also a
new `OperationRemote` type.
Secondly, operations are only created in the tracker and can only be removed
by their creators (they can be overwritten by other operations though).
Operations cannot be accessed directly and modifications are limited to setting
Error for PhaseDone operations.
After created, *Operations are queued in the pinWorker queues which handle any
status updates. This means, that, even when an operation has been removed from
the tracker, status updates will not interfere with any other newer operations.
In the maptracker, only the Unpin worker Cleans operations once processed. A
sucessful unpin is the only way that a delete() happens in the tracker map.
Otherwise, operations stay there until a newer operation for the Cid arrives
and 1) cancels the existing one 2) takes its place. The tracker refuses to
create a new operation if a similar "ongoing" operation of the same type
exists.
The final change is that Recover and RecoverAll() are not async and play by the
same rules as Track() and Untrack(), queueing the items to be recovered.
Note: for stateless pintracker, the tracker will need to Clean() operation
of type OperationPin as well, and complement the Status reported
by the tracker with those coming from IPFS.
License: MIT
Signed-off-by: Hector Sanjuan <code@hector.link>
2018-05-25 16:32:10 +00:00
|
|
|
|
|
|
|
|
// MustDecodeCid provides a test helper that ignores
|
|
|
|
|
// errors from cid.Decode.
|
2018-09-22 01:00:10 +00:00
|
|
|
func MustDecodeCid(v string) cid.Cid {
|
Fix: maptracker race issues
This commit attempts to fix race issues in the maptracker since the
introduction of the OperationTracker.
There were two main problems:
* Duplicity tracking the state both in the state map and the opTracker
* Non atomiciy of operations with different threads being able to affect
other threads operations.
A test performing random Track/Untracks on the same Cid quickly showed
that items would sometimes stay as pin_queued or pin_unqueued. That happened
because operations could be cancelled under the hood by a different request,
while leaving the map status untouched.
It was not simply to deal with this issues without a refactoring.
First, the state map has been removed, and the operation tracker now provides
status information for any Cid. This implies that the tracker keeps all
operations and operations have a `PhaseDone`. There's also a
new `OperationRemote` type.
Secondly, operations are only created in the tracker and can only be removed
by their creators (they can be overwritten by other operations though).
Operations cannot be accessed directly and modifications are limited to setting
Error for PhaseDone operations.
After created, *Operations are queued in the pinWorker queues which handle any
status updates. This means, that, even when an operation has been removed from
the tracker, status updates will not interfere with any other newer operations.
In the maptracker, only the Unpin worker Cleans operations once processed. A
sucessful unpin is the only way that a delete() happens in the tracker map.
Otherwise, operations stay there until a newer operation for the Cid arrives
and 1) cancels the existing one 2) takes its place. The tracker refuses to
create a new operation if a similar "ongoing" operation of the same type
exists.
The final change is that Recover and RecoverAll() are not async and play by the
same rules as Track() and Untrack(), queueing the items to be recovered.
Note: for stateless pintracker, the tracker will need to Clean() operation
of type OperationPin as well, and complement the Status reported
by the tracker with those coming from IPFS.
License: MIT
Signed-off-by: Hector Sanjuan <code@hector.link>
2018-05-25 16:32:10 +00:00
|
|
|
c, _ := cid.Decode(v)
|
|
|
|
|
return c
|
|
|
|
|
}
|
