The built-in receiver
The manager ships its own syslog listener — UDP and TCP, both on port 5514, both enabled by default. Messages are stored and streamed live to the Logs page as they arrive.
TCP is worth choosing when UDP is filtered somewhere in the path, or when you would rather have delivery than speed; RouterOS 7 supports both.
Configuring a device
Nothing is collected until a router is told to send. The Logs page has a Setup Guide that generates these commands and can save them as a Command Template, so there is nothing to retype. It fills in the address you opened the manager on:
/system logging action add name=mgr target=remote remote=<MANAGER_IP> remote-port=5514
/system logging add topics=info action=mgr
/system logging add topics=warning action=mgr
/system logging add topics=error action=mgr
/system logging add topics=critical action=mgr
When the manager runs as a container on a MikroTik router, the address you opened it on is usually the router's, and logs sent there never reach the container. Use the container's own address, the one on its veth interface. The Setup Guide says this under the address field and fills nothing in for you, because in a plain Docker install the container's address is not reachable from other routers. (v1.107.0+)
For TCP instead of UDP the action carries remote-protocol=tcp. Which topics you send is your decision — the four above are a reasonable default, and adding more means more volume to store.
Messages appear as they do in the device's own log. The receiver also accepts the RFC3164-style forms, so a device configured the BSD way still lands correctly.
Reading the page
Every line names the router it came from, in its own column between the time and the topics. The name links to that device, and hovering the line spells it out in full together with the exact time and the severity — worth knowing when a fleet's names all begin the same way and the column has to trim them. A line the manager could not attribute reads Unknown with the address it arrived from.
The filters above the stream narrow it by site, by device, by severity, by topic, by text in the message, and by time — either one of the preset ranges or a from/to of your own. They combine, and they apply to entries arriving live as well as to what is already on screen, so a view narrowed to one site stays that way as new lines come in. On an install with a long device list, the box beside the device selector narrows it by name or address, and choosing a site narrows it to that site's routers on its own.
The filters are kept in the page address, so going back to the page from a device keeps them, and the address can be shared or bookmarked with the filter in it. Opening Logs from the menu starts with no filter. (v1.107.0+)
Pause holds the view still while you read something, counting what arrives meanwhile and releasing it when you resume. Auto-scroll keeps you at the newest entry and switches itself off when you scroll away to read, back on when you return to the top. Load older at the foot of the list walks backwards through what is stored.
When a filter matches nothing, the page says so. That is a different message from the empty page you get before any device has been configured to send, which offers the setup guide instead.
Analysis
Analysis at the top of the Logs page shows what changed in your logs over the last 24 hours, so you don't have to read the stream to find it. Each device is compared with its own usual day, not with the rest of the fleet. A site and a device selector narrow the page. (v2.0.0+)
The findings come in groups. A group with nothing in it is left out.
| Group | What it lists |
|---|---|
| New | Lines the device has not sent in the 30 days before. |
| Changed | Lines sent far more often than on a usual day for that device. |
| Shifted | A line that, from a given hour, comes at least three times more often, or stopped coming. |
| Log level moved | Everything the device logs moved to a new level, and from when. |
| Unusual hour | An hour whose volume and mix of lines looks like none of the device's other hours. |
| Gone quiet | Devices that always log, sent nothing for 6 hours, and are online. |
| Ongoing | Warnings and errors that were there on several days and still are. |
Every finding links to the lines it came from. A device with fewer than seven days of logs is still learning its usual day, and the page counts how many are.
Known problems
Some problems look normal to that comparison because they have been there since the device started logging. A tunnel that fails every ten seconds for a week is that device's usual day. Known problems, at the top of the page, names these from a device's first day, from its last 24 hours of lines. (v2.1.0+)
| Problem | When it is listed |
|---|---|
| Password guessing | 5 failed sign-ins from one address within 10 minutes. |
| A tunnel that does not stay up | SSTP and the other PPP tunnels: 6 disconnects, with the last reason the router gave. WireGuard: 6 handshakes that did not complete. IPsec: the same error 6 times. |
| A flapping port | 4 link downs of one port within an hour. The port is shown with its interface comment, if it has one, so you can see what is connected to it. (v2.3.0+) |
| SYN flood | Any possible SYN flooding line. |
| IP address conflict | A DHCP client turned the same address down 3 times, which means something else already uses it. |
| A script that keeps failing | 3 failures of one script in 24 hours, named by the scheduler entry or session it ran from. |
| A BGP session that keeps dropping | The session went Idle 3 times. |
Each of these problems is also listed in Needs attention on Overview, in amber, where it can be acknowledged like any other row. The list there is refreshed once a minute. (v2.3.0+)
Changes and sign-ins
A section of its own lists two things about people rather than faults. They stay on this page and do not go to Needs attention. (v2.1.0+)
- Configuration changed outside the manager. A change made in WinBox or the router's web interface, or over SSH or the API by an account other than the one the manager uses for that device. It shows the last five commands. A person who signs in over SSH with the manager's own account looks the same as the manager, so that is not listed.
- A sign-in from somewhere new. A user, address and method the device has not seen before. It starts once the device has three days of sign-in history, so the first days do not list every usual sign-in. API and REST sign-ins are left out, as they are usually automation.
Wi-Fi clients
Below the findings, the page reads the connect and disconnect lines your access points log and points out the clients that stand out: ones that drop at an even rhythm, drop on several access points, drop far more often than the rest, have a weak signal, bounce between two radios within minutes, or get addresses on more than one network. With the older CAPsMAN it also shows clients the access list keeps refusing. An access point where many clients drop is named above the list. The window can be the last 6 hours, 24 hours, 3 days or 7 days, and Show all clients lists everyone, not only those that stand out.
This needs the access points to send their wireless lines to the receiver. Without them the section says there is nothing to read.
A device's own Logs tab starts with the same summary for that one device.
How a log finds its device
Incoming messages are matched to a device by any of its interface addresses, collected on each poll — not only by the address you manage it on. That is why a router with several addresses does not need its source address pinned for logs to be attributed correctly.
A device you added by a name rather than an address — an IP/Cloud name, or any other dynamic-DNS name — has that name resolved, and the address it points to is matched as well. A changed address is picked up within two minutes. Where two devices resolve to the same address, which is what happens when several routers sit behind one NAT, their lines are deliberately left unattributed: a log line labelled with the wrong router is worse than one labelled with none.
If nothing matches on address, the manager falls back to the hostname in the message, comparing it against the device's name and its RouterOS identity. That field only exists when the logging action is set to bsd-syslog=yes; with the default format there is no hostname to read.
If entries show up as unknown with a bare IP just after setup, the usual cause is that the manager has not polled that device since, so it does not yet know that address belongs to it. Waiting for a poll normally fixes it; pinning the source address on the logging action fixes it immediately.
Retention
Logs are pruned on two axes so that neither a long quiet period nor one very chatty device fills the database: by age, seven days by default, and by rows per device, ten thousand by default. Both are adjustable in Settings, and either can be overridden for a single device on its own form — useful when one router matters more, or talks far more, than the rest.
Scoped users see the logs of their own sites only, the same as everywhere else.
Exporting logs
To hand logs to someone else, such as your provider or a colleague, set the filters on the Logs page or on a device's Logs tab so they show what you need, then press Export. You can choose between two formats:
- CSV opens in Excel or LibreOffice, one column per field, with both local time and UTC.
- Plain text puts one entry per line, like the device's own log. A short header at the top records the time zone and the filter that was used, so it can be pasted into an email as it is.
The file holds everything the filter matches, not only the rows on screen, oldest first and in your browser's time zone. One export stops at the newest 100,000 entries. When there are more, the page says so, and a narrower time range gets you the older ones.
An export follows the same site access as the page, so a scoped user exports only the logs of their own sites.
For a few lines there is Copy next to Export. It copies only the lines on screen, as plain text with the date, time, topics and message (and the device on the Logs page), so you can paste them straight into a chat or a document. Lines loaded with Load older are included.
Forwarding to other collectors
The manager can pass every line it receives on to other syslog servers, such as Wazuh, Graylog or a backup syslog. Lines are still stored here as before. It is off by default. Turn it on in Settings → Syslog with Forward to other collectors, then list the servers under Collectors, up to eight. (v2.0.0+, several collectors v2.2.0+)
| Column | What it does |
|---|---|
| Address | An IPv4 or IPv6 address, or a DNS name. |
| UDP port | 514 by default. |
| From level | Only lines at this level or more severe go to this collector. debug, the default, forwards everything. |
| On | Switches one collector off without removing it. |
Each collector has its own level, so Wazuh can get everything while a backup syslog keeps only errors. Add collector adds a row and Save stores the list. One collector that is down or cannot be resolved does not stop the others. An install that had a single collector before v2.2.0 shows it as the first row.
Lines go out over UDP in the RFC 3164 format. Every packet the collector gets comes from the manager's address, so each line carries the router's name in the manager as its host name (a space in the name becomes a dash). A line from a source the manager does not know carries that source's address instead. After the host name comes the line as the router logged it, topics included. The router's own timestamp goes with it when the router sent one.
If a collector's address and port point back at the manager's own receiver, nothing is sent to it, because every line would come back in and go out again. The manager notes this in its own log.
For Wazuh there are ready-made decoders, rules and a dashboard, see Wazuh. For Graylog there is a content pack with a stream and a dashboard, see Graylog.
Log alerts
A rule watches the lines your devices send and fires the Log Alert webhook when one matches. Rules live under Logs → Alerts. To be told anything, subscribe a webhook to Log Alert on the Webhooks page. (v1.102.0+)
A rule has these parts:
| Field | What it does |
|---|---|
| Text to match | Contains, or a regular expression in JavaScript syntax. Both ignore case. Up to 200 characters. |
| Topics | Optional. RouterOS topics such as system,account, separated by commas. Every topic listed has to be on the line. |
| Severity | Any, or a level and everything more serious: error, warning, notice or info. With a level set, a line that carries no severity does not match. |
| Where | All devices, one site, one tag or one device. |
| Alert after / Within | How many matching lines, inside how many minutes. The default, 1, fires on the first line. |
| Then stay quiet | Minutes after an alert during which the rule says nothing more about that device. At least 1, 15 by default. |
Counting is per device. Five within ten minutes means five matching lines from the same router, not five across the fleet, and the quiet period is per device too.
A line from an address the manager does not know as a device only reaches rules set to All devices, because there is no site, tag or device to check it against. Its alert goes only to webhooks that are not limited to sites or tags.
Test against last 24h in the rule form runs the rule as it stands, before you save it, over the lines stored in the last day. It says how many lines matched, on how many devices, and how many alerts the rule would have sent with its count and quiet period, with up to 20 example lines. It reads at most the newest 50,000 lines and says so when the day held more. Nothing is saved and nothing is sent.
Limits
Rules run on every line as it arrives, so their cost is capped. A line is cut to 2,048 characters before any rule sees it. A regular expression that could stall the receiver is refused when you save it: nested repetition such as (a+)+, backreferences, or a pattern that runs too slowly on a long line. Across all rules, at most 30 log alerts go out per minute. The rest are not sent and are counted in the manager's own log.
Only a superadmin or an administrator with no site limits can see and change rules, because a rule reads every device's lines. Creating, changing and deleting a rule is recorded in the activity log. Rules need the receiver: with it switched off, the page says so and no rule can match.
What else the log stream feeds
The log receiver is not only a viewer. Repeated failed logins seen in this stream are what the auto-block feature counts, aggregated per source address across the whole fleet — so switching on log collection is also what makes that feature possible. It is opt-in per device and has an audit-only mode; the Security page is where it lives.
It also feeds the log alerts described above.