dendi-wang/thingsboard
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Refactoring

Browse Source
This commit is contained in:
Yevhenii 2025-07-29 18:36:07 +03:00
parent a101c9403f
commit 2c7c63c537
2 changed files with 7 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
application/src/main/java/org/thingsboard/server/service/edge/rpc/fetch/GeneralEdgeEventFetcher.java
View File

@ -36,7 +36,9 @@ public class GeneralEdgeEventFetcher implements EdgeEventFetcher {
private boolean seqIdNewCycleStarted;
private Long maxReadRecordsCount;
private final EdgeEventService edgeEventService;
// Subtract from queueStartTs to ensure no data is lost due to potential misordering of edge events by created_time.
// Subtract from queueStartTs to compensate for possible misalignment between `created_time` and `seqId`.
// This ensures early events with lower seqId are not skipped due to partitioning by `created_time`.
// See: edge_event is partitioned by created_time but sorted by seqId during retrieval.
private final long misorderingCompensationMillis;
@Override

6
application/src/main/resources/thingsboard.yml
View File

@ -1480,8 +1480,10 @@ edges:
no_read_records_sleep: "${EDGES_NO_READ_RECORDS_SLEEP:1000}"
# Number of milliseconds to wait before resending failed batch of edge events to edge
sleep_between_batches: "${EDGES_SLEEP_BETWEEN_BATCHES:60000}"
# Number of milliseconds to subtract from queue start timestamp to compensate for potential
# misordering of edge events by created_time. Prevents skipping early events.
# Time (in milliseconds) to subtract from the start timestamp when fetching edge events.
# This compensates for possible misordering between `created_time` (used for partitioning)
# and `seqId` (used for sorting). Without this, events with smaller seqId but larger created_time
# might be skipped, especially across partition boundaries.
misordering_compensation_millis: "${EDGES_MISORDERING_COMPENSATION_MILLIS:3600000}"
# Max number of high priority edge events per edge session. No persistence - stored in memory
max_high_priority_queue_size_per_session: "${EDGES_MAX_HIGH_PRIORITY_QUEUE_SIZE_PER_SESSION:10000}"