View Issue Details

IDProjectCategoryView StatusLast Update
0025450Return of ReckoningGeneralpublic2026-05-17 10:06
ReporterWholdar Assigned To 
PrioritynormalSeverityminorReproducibilityalways
Status newResolutionopen 
Summary0025450: GraphQL API kill/scenario feeds stale after 2026-05-14 patch/restart
DescriptionAfter the patch/restart this morning ("Up since: 2026-05-14 06:22:51"), the public GraphQL API appears to serve no new kills or completed scenarios.

Checked at: 2026-05-14 18:23 UTC
Endpoint: https://production-api.waremu.com/graphql

The API responds with HTTP 200, but newest records are still from before/around the restart window.

Observed:
- Newest kill returned is 2026-05-14T05:53:48Z.
- Newest scenario returned ended 2026-05-14T05:15:27Z.
- Filtering for records after 2026-05-14T06:22:51Z returns no kills and no scenarios.

Expected:
Kills and completed scenarios after the patch/restart should be present.
Steps To ReproduceRun this GraphQL request:

curl -sS -X POST https://production-api.waremu.com/graphql \
  -H 'content-type: application/json' \
  -H 'cache-control: no-cache' \
  --data-raw '{"query":"query ApiFreshness { latestKills: kills(first: 3, includeAssists: false) { nodes { id time scenario { id name } position { zone { id name } } } } killsAfterRestart: kills(first: 3, where: { time: { gte: \"2026-05-14T06:22:51Z\" } }, includeAssists: false) { nodes { id time } } latestScenarios: scenarios(first: 3) { edges { node { id startTime endTime scenario { id name } tier queueType } } } scenariosAfterRestart: scenarios(first: 3, where: { endTime: { gte: \"2026-05-14T06:22:51Z\" } }) { totalCount edges { node { id endTime scenario { id name } } } } }"}'

Additional InformationActual output from the API:

{
  "data": {
    "latestKills": {
      "nodes": [
        {
          "id": "64975800",
          "time": "2026-05-14T05:53:48.673271Z",
          "scenario": null,
          "position": {
            "zone": {
              "id": "102",
              "name": "High Pass"
            }
          }
        },
        {
          "id": "64975799",
          "time": "2026-05-14T05:53:36.942404Z",
          "scenario": null,
          "position": {
            "zone": {
              "id": "102",
              "name": "High Pass"
            }
          }
        },
        {
          "id": "64975798",
          "time": "2026-05-14T05:53:24.716955Z",
          "scenario": null,
          "position": {
            "zone": {
              "id": "102",
              "name": "High Pass"
            }
          }
        }
      ]
    },
    "killsAfterRestart": {
      "nodes": []
    },
    "latestScenarios": {
      "edges": [
        {
          "node": {
            "id": "da55da80-a350-4cbd-b7db-145781b76c55",
            "startTime": "2026-05-14T05:03:38.88645Z",
            "endTime": "2026-05-14T05:15:27.681468Z",
            "scenario": {
              "id": "2109",
              "name": "Reikland Hills"
            },
            "tier": 3,
            "queueType": 0
          }
        },
        {
          "node": {
            "id": "dfa1ccbb-b9b2-462c-be52-c09468016e49",
            "startTime": "2026-05-14T03:02:04.239928Z",
            "endTime": "2026-05-14T03:12:02.995434Z",
            "scenario": {
              "id": "2206",
              "name": "Blood of the Black Cairn"
            },
            "tier": 3,
            "queueType": 0
          }
        },
        {
          "node": {
            "id": "7ccd8330-93c6-4635-9b75-2708edf6cb1d",
            "startTime": "2026-05-14T02:36:41.612797Z",
            "endTime": "2026-05-14T02:46:15.573771Z",
            "scenario": {
              "id": "2109",
              "name": "Reikland Hills"
            },
            "tier": 1,
            "queueType": 0
          }
        }
      ]
    },
    "scenariosAfterRestart": {
      "totalCount": 0,
      "edges": []
    }
  }
}
TagsNo tags attached.

Activities

Wholdar

Wholdar

2026-05-17 10:06

reporter   ~0048393

Scenarios and Lost Vale data look like they are working again now. The remaining
issue I can still reproduce is with the public GraphQL `kills` connection.

Checked at: 2026-05-17 09:50 UTC
Endpoint: https://production-api.waremu.com/graphql

Current observed behavior:
- `scenarios` and Lost Vale are not the stale-feed issue anymore.
- `kills(first: ...)` still returns old kill rows from 2026-05-15 instead of
  current 2026-05-17 rows.
- Filtering `kills` for rows on/after 2026-05-16 returns an empty list.
- Some valid `kills` pagination/filter queries also hit backend MySQL errors.
  I saw "The Command Timeout expired before the operation completed." Earlier in
  the same check, `kills(last: 5)` returned "Packet received out-of-order.
  Expected 1; got 2."

Reproduction 1 - latest kills are stale

curl -sS --get \
  --data-urlencode 'query=query ($first: Int) { kills(first: $first, includeAssists: false) { nodes { id time position { zone { id name } } scenario { id name } } } }' \
  --data-urlencode 'variables={"first":3}' \
  https://production-api.waremu.com/graphql

Observed output excerpt:

{
  "data": {
    "kills": {
      "nodes": [
        {
          "id": "65009644",
          "time": "2026-05-15T00:43:49.8554Z",
          "position": {
            "zone": {
              "id": "236",
              "name": "Lost Temple of Isha"
            }
          },
          "scenario": {
            "id": "2203",
            "name": "Lost Temple of Isha"
          }
        }
      ]
    }
  }
}

Reproduction 2 - fresh kill window is empty

curl -sS --get \
  --data-urlencode 'query=query ($first: Int, $where: KillFilterInput) { kills(first: $first, where: $where, includeAssists: false) { nodes { id time } } }' \
  --data-urlencode 'variables={"first":1,"where":{"time":{"gte":"2026-05-16T00:00:00Z"}}}' \
  https://production-api.waremu.com/graphql

Observed output:

{
  "data": {
    "kills": {
      "nodes": []
    }
  }
}

Reproduction 3 - valid kills pagination can throw backend errors

curl -sS --get \
  --data-urlencode 'query=query ($last: Int) { kills(last: $last, includeAssists: false) { nodes { id time } } }' \
  --data-urlencode 'variables={"last":1}' \
  https://production-api.waremu.com/graphql

Observed output excerpt:

{
  "errors": [
    {
      "message": "Unexpected Execution Error",
      "path": [
        "kills"
      ],
      "extensions": {
        "exception": {
          "message": "The Command Timeout expired before the operation completed."
        }
      }
    }
  ],
  "data": {
    "kills": null
  }
}

Expected behavior:
`kills(first: ...)` or a time-filtered `kills(where: { time: { gte: ... } })`
should return recent kill rows when there has been live kill activity. Valid kill
pagination should also not fail with MySQL connector / command timeout errors.

Issue History

Date Modified Username Field Change
2026-05-14 18:28 Wholdar New Issue
2026-05-17 10:06 Wholdar Note Added: 0048393