Tally Connect
Operations

Daily checks and maintenance

A practical routine for checking sync health, handling interruptions, and resuming after maintenance.

Use this routine whenever the team starts work and after changes to the Windows computer, Tally company, or connector configuration.

Start-of-day checks

  1. Open TallyPrime under the Windows user who owns the Agent installation and load the bound company.
  2. In Tally Connect Home or Setup, confirm a recent heartbeat and the expected company. Check the Agent version if an update was performed.
  3. Open Activity and review delayed, failed, and blocked work.
  4. Check Orders for recent verified Sales Vouchers and refund Credit Notes.
  5. Review open inventory conflicts and paused baselines for participating items.

An online Agent does not establish that all workflows are ready. Company access, master evidence, mappings, and write permissions are checked separately.

Decide what needs attention

ObservationNext actionRecovery evidence
Agent offlineRestore the scheduled Agent and run diagnosticsRecent heartbeat and successful Tally read
Company mismatchLoad the bound companySuccessful discovery or workflow read
Queued work while Agent is onlineInspect Activity; operator checks worker and queueProgress advances and work reaches a terminal result
Failed readCorrect the cause and use Retry readNew successful read
Uncertain voucher writeFollow voucher reconciliationVerified existing link or a recorded absence followed by a fresh verified export
Stalled product scanReview pinned evidence and recovery eligibilityAdvancing progress and reviewed outcomes
Inventory conflictReview both quantities and resolve in InventoryResolution completes and the pair is eligible again

Planned maintenance

Before updating Windows, TallyPrime, or the Agent, pause automatic order export and inventory. Let active work finish where possible and record anything that remains unresolved. Back up the Tally company using your established process.

Perform the maintenance, then reopen TallyPrime and the original company. Run Agent diagnostics, confirm the heartbeat and version, and refresh masters if company or master data changed.

Recheck readiness and mappings before resuming. Resume one workflow at a time and verify its first result in Activity and the destination system.

Unplanned interruption

Restore the same Windows user, Agent, and company. Durable queued work remains subject to current lease and safety checks. Inspect outcomes before submitting another export: an interrupted write can require reconciliation even though the computer is healthy again.

Do not clear the local outbox, remove command records, or change company binding to force work through. Preserve the evidence needed to distinguish pending work from an import that already committed.

Handoff to another operator

Record the selected company, automation settings, latest verified order/refund, open conflicts, unresolved command or scan IDs, and changes made during the shift. Include timestamps with timezone so the next operator can match the Activity log.

Useful follow-ups: Voucher recovery, Catalog scan recovery, and Inventory conflicts.