Budibase/budibase · error · HTTPError

Unable to allocate Teams app package version

Error message

Unable to allocate Teams app package version

What it means

After a semver bump, the function attempts to persist the new app package version by calling sdk.ai.agents.update inside a retry loop over deployment targets. If every attempt fails (caught errors lead to 'continue') and the loop is exhausted, the function throws this 409 HTTPError indicating it could not allocate a Teams app package version.

Source

Thrown at packages/server/src/api/controllers/ai/agents.ts:311

        ...agent,
        MSTeamsIntegration: {
          ...agent.MSTeamsIntegration,
          appPackageVersion,
        },
      })
      return { agent: updatedAgent, messagingEndpointUrl, appPackageVersion }
    } catch (error) {
      if (
        db.isDocumentConflictError(error) &&
        attempt < TEAMS_APP_PACKAGE_VERSION_RETRIES - 1
      ) {
        continue
      }
      throw error
    }
  }

  throw new HTTPError("Unable to allocate Teams app package version", 409)
}

export async function fetchTools(ctx: UserCtx<void, ToolMetadata[]>) {
  ctx.body = await sdk.ai.agents.getAvailableToolsMetadata()
}

export async function fetchAgents(ctx: UserCtx<void, FetchAgentsResponse>) {
  const agents = await sdk.ai.agents.fetch()
  ctx.body = { agents: agents.map(toAgentResponse) }
}

export async function createAgent(
  ctx: UserCtx<CreateAgentRequest, CreateAgentResponse>
) {
  const body = ctx.request.body
  const createdBy = ctx.user?._id!
  const globalId = db.getGlobalIDFromUserMetadataID(createdBy)
  const projectIds = await resolveProjectIds(body.projectIds)

View on GitHub (pinned to a81a902e9a)

Solutions

1. Retry the operation — the 409 signals a conflict, likely a concurrent deployment of the same agent
2. Ensure only one deployment per agent runs at a time (serialize at the caller, add a lock)
3. Check server logs for the underlying error from sdk.ai.agents.update that caused every iteration to fail
4. Verify the agent still exists and MSTeamsIntegration is intact, then redeploy

// Ensure no concurrent deployment is in flight for this agent before calling
const lockKey = `teams-deploy:${agent._id}`
if (await cache.get(lockKey)) throw new Error("A Teams deployment is already in progress for this agent")
await cache.store(lockKey, true, { ttl: 60 })

try {
  await allocateMSTeamsAppPackageVersion({ agent, messagingEndpointUrl })
} catch (err) {
  if (err instanceof HTTPError && err.status === 409) {
    // wait for the competing deployment to finish, then retry once
    await new Promise(r => setTimeout(r, 2000))
    await allocateMSTeamsAppPackageVersion({ agent, messagingEndpointUrl })
  } else { throw err }
}

When it happens

Trigger: All iterations of the update loop fail — e.g. concurrent updates from multiple calls racing on the same agent, sdk.ai.agents.update throwing repeatedly (conflict/optimistic-locking or API failure), then control falls past the loop to the final throw.

Common situations: Two users/automations deploying the Teams app for the same agent simultaneously; transient CouchDB write conflicts during every retry; agent document deleted mid-loop.

Related errors

• Invalid Teams app package version
• Failed to clear project assignments.
• Unable to bulk remove documents: ${res.error}
• Unable to remove top level directory - some skeleton files a
• Group name "${name}" is unavailable

AI-assisted analysis of Budibase/budibase@a81a902e9a (2026-08-29). Data as JSON: /api/errors/d0da3ef50319d2ba. Report an issue: GitHub.