This is due to the mirroring functionality trying to skip messages when it detects a gap. In a replicated stream this puts excessive stress on the raft system.
This step is not needed at all if the mirror stream has no messages, we can simply jump ahead.
Signed-off-by: Derek Collison <derek@nats.io>
Bail early if new consumer, meaning stream sequence floor is 0.
Decide which linear space to scan.
Do no work if no pending and we just need to adjust which we do at the end.
Also realized some tests were named wrong and were not being run, or were in wrong file.
Signed-off-by: Derek Collison <derek@nats.io>
Under asymmetric network latency based clusters, if a node in an R3 was replicating a consumer and the parent stream, but was the leader of neither, but the path from the stream leader was faster then the consumer leader a replicated ack could arrive before the message itself.
In this case we used to forward a delete message request to the stream leader which would then replicate that to all stream replicas, causing more work which could lead to increased publisher times on clients connected to the slow node.
Signed-off-by: Derek Collison <derek@nats.io>
This could lead to instability in the system.
The bug would manifest in replicated consumers when certain messages could be acked out of order, and, the pending list would never go to zero.
Signed-off-by: Derek Collison <derek@nats.io>
This can help sync on restarts and improve ghost ephemerals. Also added more code to suppress respnses and API audits when we know we are recovering.
Signed-off-by: Derek Collison <derek@nats.io>
In cases where we had a large subject space, a filestore with many msg blocks, and a filtered consumer with a wildcard filtered subject, creating a consumer could take more memory and time then we wanted.
This improvement works when the consumer is DeliverAll and we used the upper layer in memory psim structure to scan but only in memory and avoid a file read for each msg block.
Signed-off-by: Derek Collison <derek@nats.io>
We noticed this was being called alot in user environments.
When the consumer was filtered with a wilcard and the stream had a high cardinality of subjects and was falling behind this could take a substantial amount of time.
Signed-off-by: Derek Collison <derek@nats.io>
When a stream had a large number of consumers on a server that were sparse, the signaling mechanism would do a linear scan to signal matching consumers. As usage patterns have continued to have more consumers that are filteres and sparse, meaning a message is destined for a single or small number of consumers.
This change moves selection to a sublist that tracks only active consumer leaders for selection, which optimizes selection of consumers to signal when the number of consumers is large.
Signed-off-by: Derek Collison <derek@nats.io>
The bug was when a timestamp for the pending state was exactly -1 which could happen based on timing of the redlivered pending items which would set pending.Timestamp into the future potentially and the timing on the encodeConsumerState call.
Minor fixes to raft.
Signed-off-by: Derek Collison <derek@nats.io>
When we deleted a consumer from an interest policy stream we would make sure to clean up any unacked messages.
However we only based start from the ack floor for the consumer and did not take into account the first sequence of the stream.
Signed-off-by: Derek Collison <derek@nats.io>
A request to `$SYS.REQ.SERVER.PING.JSZ` would now return something
like this:
```
...
"meta_cluster": {
"name": "local",
"leader": "A",
"peer": "NUmM6cRx",
"replicas": [
{
"name": "B",
"current": true,
"active": 690369000,
"peer": "b2oh2L6w"
},
{
"name": "Server name unknown at this time (peerID: jZ6RvVRH)",
"current": false,
"offline": true,
"active": 0,
"peer": "jZ6RvVRH"
}
],
"cluster_size": 3
}
```
Note the "peer" field following the "leader" field that contains
the server name. The new field is the node ID, which is a hash of
the server name.
Signed-off-by: Ivan Kozlovic <ivan@synadia.com>
