What causes "no sessions for workspace" in hcengineering/platform?

No fix required if shutdown is expected - this is informational logging at warn level.

no sessions for workspace

Error message

no sessions for workspace

What it means

TSessionManager logs 'no sessions for workspace' during performWorkspaceCloseCheck when a tracked workspace's session set has dropped to zero. It is the first step of the automatic idle-workspace teardown: if no sessions remain, the manager closes the workspace, removes it from the workspaces map, and emits a workspace 'down' event. It is a diagnostic warning, not a failure - it tells you why a workspace is being shut down.

Source

Thrown at foundations/server/packages/server/src/sessionManager.ts:1271

      compression
    )
  }

  async closeWorkspaces (ctx: MeasureContext): Promise<void> {
    clearInterval(this.checkInterval)
    for (const w of this.workspaces) {
      await this.doCloseAll(w[1], 1, 'shutdown')
    }
    await this.workspaceProducer.close()
    await this.usersProducer.close()
  }

  private async performWorkspaceCloseCheck (workspace: Workspace): Promise<void> {
    const uuid = workspace.wsId.uuid
    const logParams = { uuid, url: workspace.wsId.url }
    if (workspace.sessions.size === 0) {
      if (LOGGING_ENABLED) {
        this.ctx.warn('no sessions for workspace', logParams)
      }
      try {
        if (workspace.sessions.size === 0) {
          await workspace.close(this.ctx)

          this.workspaces.delete(uuid)
          workspace.context.end()
          if (LOGGING_ENABLED) {
            this.ctx.warn('Closed workspace', logParams)
          }

          await this.workspaceProducer.send(this.ctx, workspace.wsId.uuid, [workspaceEvents.down()])
        }
      } catch (err: any) {
        Analytics.handleError(err)
        this.workspaces.delete(uuid)
        if (LOGGING_ENABLED) {
          this.ctx.error('failed', { ...logParams, error: err })

View on GitHub (pinned to 63e28dc964)

Solutions

No fix required if shutdown is expected - this is informational logging at warn level.
If the workspace should stay alive, ensure at least one session remains connected (keep the client attached) before the close check runs.
If the workspace is closing unexpectedly, check session disconnect/timeout logic that removes sessions from workspace.sessions.
Inspect workspaceProducer.send(workspaceEvents.down()) consumers to confirm they handle workspace down events.

Defensive patterns

Strategy: validation

Validate before calling

if (workspace.sessions.size > 0) {
  // keep workspace alive; skip close check
}

Type guard

function hasSessions (ws: { sessions: Set<unknown> }): boolean {
  return ws.sessions.size > 0
}

Prevention

Keep at least one session attached if the workspace must stay up.
Treat the log as informational; alert on repeated unexpected occurrences instead.
Monitor workspace down events from workspaceProducer to correlate client disconnects.

When it happens

Trigger: The periodic/deferred close check runs for a workspace after its last client session has disconnected or been removed, so workspace.sessions.size === 0 when performWorkspaceCloseCheck executes.

Common situations: Users closing all browser tabs attached to a workspace; sessions dropped due to network loss; session cleanup after client disconnect; normal workspace idle shutdown observed in server logs while investigating why a workspace went down.

Related errors

AI-assisted analysis of hcengineering/platform@63e28dc964 (2026-08-29). Data as JSON: /api/errors/5994df653ce86225. Report an issue: GitHub.