Elbuz logs every scheduled run and every refusal with its reason in the “Operation log”, so finding why the night was quiet takes minutes.
When you need this:
- No log entries all night — six checks, from the most common cause to rare ones.
- A source imports manually, not on schedule — a disabled source or supplier is skipped with a log warning.
- A “data processing conflict” entry — the run overlapped another and was skipped until the next match.
- A run broke off, the schedule is silent — the app releases a stuck lock itself within 10–60 minutes.
Where to start: open the “Operation log” for that night.
What helps you find the cause
- Six checks in order — from the most common cause to the rarest; each one makes sense only once the previous one has passed.
- The “Source … was skipped by the schedule” warning — says why the schedule didn't pick up a source that imports fine with the button.
- The “data processing conflict” entry — shows that the run overlapped with another run of the same kind.
- Stuck-lock watchdogs — clear what a broken-off run left behind, so the schedule works again.
- Message table — every schedule refusal with its cause and what to do.
Six checks when nothing happened
The order isn't random: the checks go from the most common cause to the rarest, and each one makes sense only after the previous one has passed.
- Action or result
- Decision: the path depends on the answer
The schedule is silent: what to check
Check one: the log. Empty for the whole night — the run either never started, or started and exited at once without leaving a trace; both cases are covered below. There's a refusal entry — the run did start, and the reason is written right in it; see the next section.
An empty log doesn't prove the app did nothing. Some things pass without an entry: a parser's own schedule when the parsing module is disabled, an empty job queue task, and a queue task skipped because its template is busy.
Check two: the “Active” checkbox. The most common cause. The app ignores a disabled schedule, which looks like a failure: an empty log, no messages. Also check that the builder's rule reads as words: an empty rule is skipped silently too. Why the checkbox can be cleared with a rule set — see the schedule window.
Check three: the item and whoever it belongs to. The schedule skips a disabled source, a disabled supplier, and a browser-based , even though the button imports them. How to recognize this — in the next section.
Check four: the time zone. Compare the time with the Time zone field in the schedule window: “10:00” means 10:00 in the time zone of the admin user. If the field is empty, or clearly shows a different time zone from yours, contact support.
Check five: a refusal because of another run. A conflict is always logged, so that's exactly where to look for it — an entry with the words “data processing conflict”. If there's no such entry and still no runs, see the section on a stuck lock: it's just as silent as a cleared checkbox.
Check six: the work itself. A schedule doesn't fix anything: a template that fails when run manually will fail at night too, only nobody will be watching. But note the flip side: a successful manual run does not clear check three — the button ignores the state of the source and the supplier, while the schedule doesn't.
A seventh cause is invisible in the interface: the app's every-minute pass skips free accounts. The schedule saves, the checkbox is on — and nothing happens. A deactivated account too, but you'll notice it: a placeholder page opens instead of the app. Six checks pass, the log empty for weeks? Contact support — the interface can't fix this.
Why a source imports with the button but not on schedule
This is the cause that leaves most people stuck. The schedule starts the same work as the button, but before starting, it filters out things the button doesn't. The schedule skips a price list if:
- the source is disabled. It's visible in the price list list with the “Disabled” status badge; you enable it with “Enable source” in the row menu;
- the supplier the source belongs to is disabled. Then the source doesn't appear in the price list list at all — find the supplier and the active checkbox in its card;
- the source's parser is set up to run in the browser, not on our server. Such a parser works only while you're in the app, and no schedule will start it;
- the source has been deleted, but its schedule remains. Such an entry gets in nobody's way, but it won't do anything either — and you can't currently remove it from the interface.
In the first three cases, the “Import” button works, but the schedule doesn't. So the “does it run manually?” check misses this: manually, it runs. But each skip leaves a log warning, “Source … was skipped by the schedule”, naming the reason: the source or supplier is disabled, the source was deleted, or the parser runs in the browser.
1
Disabled pill2
Active toggle- Disabled — in this state the schedule silently skips the source.
- The toggle next to the pill — the only place where the source is switched back on.
The schedule skips a disabled export template in the same way — see the article on export templates for details.
Website is a separate case: on schedule, it exits silently if the website parsing module is disabled for the account. No log entry, no message. A price list source fed by a server-side parser is skipped then too, but with the warning “server-side parsing is not enabled for the account”.
Why a run was skipped because of another run
Before starting, each handler checks whether another one of the same kind is running. If it is, the run isn't postponed — it's skipped until the rule's next match. So a schedule set to “every day at 3:00” that runs into another run won't try again in five minutes, but tomorrow at three.
In the log, this appears as one of four entries, depending on what didn't run:
- “Scheduled export canceled because another template is being exported (data processing conflict).”
- “Scheduled price list import canceled: another price list is being loaded (data processing conflict).”
- “Scheduled import canceled due to another template being imported (data processing conflict).”
- “Scheduled download canceled because another parsing template is being loaded (data processing conflict).”
There's also a fifth entry, and it's about something else — not two different schedules, but one and the same source:
This source is already being imported (started at 03:00). Wait for it to finish or stop it with the “Stop” button.
It means the previous run of this price list is still alive — or broke off less than fifteen minutes ago. The schedule leaves this source alone and waits for the rule's next match. You can see the unfinished run on the “Log” tab of the source card.
The fix is to spread the schedules out in time. One subtlety: the “Processing price lists” schedule, which takes all sources at once, holds the same lock as a single source's schedule. So “all price lists at 2:00” and “one heavy price list at 2:00” conflict, even though they're set up in different places.
Template imports also share the lock with the “Marketplace orders” system job, which runs every 5 minutes by default. If another import is running, even a manual one, the job silently skips that pass. A template schedule that hits an orders pass waits up to 10 minutes for it; after that it's canceled with “Scheduled import canceled…”.
If imports and exports clash, the consultant chat opened from the price list or export window offers the “Schedule without conflicts” hint. Its number: jobs overlapping in the next 24 hours by average duration over 30 days, plus runs canceled in 7 days. The consultant shows what starts together, how long it takes, and when to move it.
It moves a schedule with a “Change the schedule time” plan and a confirmation button; this requires access to the schedule window — without it, the consultant only suggests a time. The consultant marks a “Read only” system job as one that can't be changed and offers to move the other, regular job in the pair.
What to do when a run lock gets stuck
A run sets the lock when it starts and releases it when it finishes. A process that broke off leaves its lock behind — and the schedule is silent not because of a conflict, but because of a dead run's trace. So the app has watchdogs that check locks every minute and release those showing no signs of life.
Their thresholds differ, because different modules write to the log at different rates:
| What was running | Since start | Log silence |
|---|---|---|
| any job | 60 min | 30 min |
| price list processing | — | 30 min |
| template-based import | 15 min | 15 min |
| website parsing, background part | — | 15 min |
| website parsing, background part that hasn't sent a single signal | 10 min | — |
| job queue | 10 min | 10 min |
| email retrieval | — | 5 min |
| search indexing | 20 min | 20 min |
Дані таблицею
| Показник | Any job | Search indexing | Template-based import | Job queue | Price list processing | Website parsing, background part | Website parsing, no signal at all | Email retrieval |
|---|---|---|---|---|---|---|---|---|
| Since start | 60 | 20 | 15 | 10 | — | — | 10 | — |
| Log silence | 30 | 20 | 15 | 10 | 30 | 15 | — | 5 |
The parsing schedule has the same thresholds as the background part of the work: if the process broke off and stopped pinging, the lock is released within 10–15 minutes, not an hour.
A watchdog doesn't stop a live run that writes to the log regularly, however long it lasts — a multi-hour import of a large price list is safe. “Regularly” means more often than every half hour: that's the allowed silence. Your worst case is an hour of waiting until the general watchdog clears the dead process's trace.
No need to wait an hour: saving the schedule form releases the lock. Open the schedule window and change anything — even a period in “Note”; the form saves itself. Open the schedule of the same item whose run broke off: only its entry's lock is released. In the log, it's the last run that started and didn't finish.
Don't confuse this with another watchdog. A price list has its own, different stuck-run check — not on a schedule, but when you open the “Log” tab of the source card, for runs older than fifteen minutes. Separate mechanisms, different thresholds: the price list log fixes its own “Running” status, while the watchdogs in this article release schedule locks.
Scenarios
The nightly run of all price lists is skipped because of a heavy source
What to do. Find the “Scheduled price list import canceled … (data processing conflict)” entry in the log, and check whether a single source's schedule is set for the same time as “Processing price lists”. Move the source's schedule an hour earlier or later.
What you get. The two runs go one after the other and no longer hold the same lock at the same time.
A supplier was disabled temporarily, but the schedule remains
What to do. Nothing: while the source or the supplier is disabled, the schedule skips them and logs the warning “Source … was skipped by the schedule”. To resume the runs, enable the source with “Enable source”, or the supplier in its card.
What you get. At the rule's next match, the source imports on its own again; you don't need to reconfigure the schedule rule.
A run broke off overnight, and you need to start it now
What to do. Find the last run in the log that started and didn't finish. Open the schedule of that exact item and change anything in it, for example “Note”.
What you get. The lock is released when the form is saved, and the rule's next match or a manual run goes through without a refusal.
What to do when the log shows a schedule refusal
| What you see | Why | What to do |
|---|---|---|
| Not a single entry in the “Operation log” for the night | The run never started. Most often, the “Active” checkbox is cleared: the schedule entry is created disabled the first time you open the window | Go through the six checks in order (checks) |
| “Scheduled export canceled because another template is being exported (data processing conflict).” | Another export was already running at that moment. The run is skipped until the rule's next match | Spread the template schedules out in time (on conflicts) |
| “Scheduled price list import canceled: another price list is being loaded (data processing conflict).” | Another price list import was running. The “all price lists” schedule holds the same lock as a single source's schedule, and it also waits for website parsing | Spread out the source schedules; check whether “Processing price lists” is set for the same time (on conflicts) |
| “Scheduled import canceled due to another template being imported (data processing conflict).” | Another template-based import was running | Spread out the import template schedules (on conflicts) |
| “Scheduled download canceled because another parsing template is being loaded (data processing conflict).” | Another website parsing run was in progress | Spread out the parser schedules (on conflicts) |
| “This source is already being imported (started at 03:00). Wait for it to finish or stop it with the ‘Stop’ button.” | The price list was already importing at that moment: the previous run is still alive or broke off less than fifteen minutes ago | Look at the “Log” tab of the source card — the unfinished run is shown there (on refusals) |
| The schedule is on, the source imports manually, but at night the log shows only the warning “Source … was skipped by the schedule” | The schedule skips a source if it's disabled, its supplier is disabled, or its parser is set up to run in the browser. The “Import” button ignores all this | Enable the source with “Enable source”; if the source isn't in the list, enable the supplier (on skipped sources) |
| The “Active” checkbox is selected, but there are still no runs | The rule may be empty or corrupted — the app skips such a job silently, without any message | Open the schedule window and rebuild the rule from scratch in the builder (on the schedule window) |
| The schedule has been on for weeks, the log is empty, all checks pass | The every-minute pass doesn't visit free or deactivated accounts. You can't see this from the app | Contact support (checks) |
FAQ
I selected the checkbox, but there are no runs. What else should I check?
Whether the item itself and its supplier are enabled (if not, the log has a “skipped by the schedule” warning), the time zone in the Time zone field, and whether the log has a “data processing conflict” entry. The order is in the six checks.
Two schedules overlapped. Will the second one run later?
No. The second run is skipped until the rule's next match, and a “data processing conflict” entry goes to the log. A schedule set to “every day at 3:00” will try again tomorrow at three. Spread heavy schedules out in time.
A run is stuck and the lock is still held. Do I have to wait an hour?
Not necessarily. Open the schedule of the item that got stuck and change anything in it — even a period in “Note”. The form saves itself, and the lock is released. Otherwise, the watchdogs release it within an hour at most.
Can a skipped run happen later the same day?
No: a skipped run waits for the rule's next match. If you need data more often, select several hours in the schedule builder, for example 3, 9, and 15.
The schedule ran, but the prices are still wrong. Where should I look?
On the “Log” tab of the source card: it shows the numbers for every import and a breakdown of the latest one. More details: price list log.
Where to find it and who has access
- Menu
- Account → Operation log — runs and refusals are shown here
- Item schedule
- source, template, or parser card → “Configure” or “Schedule setup”
- Log of a single source
- price list source card → “Log” tab
- Access right
- the “Job schedule” right in group permissions, plus the right to the item's card
All the ways to open the schedule window are in the article on setting up a schedule.
Related topics
- If you need the big picture of how to set up a schedule — scheduled runs.
- If an operation went to the background and didn't fully complete — job queue.
- If an export template isn't built on schedule — running an export.
- If the run happened but brought in the wrong data — price list log.
