Skip to content

Monitoring Logs

Incidents & Maintenance 13 min read
On this page

Logs is the account-wide record of every individual check Uptimia has run: one row per check, per probe, newest first. It answers "what did the probe in Frankfurt see at 14:31". For the outage record (when an incident opened, how long it ran, who was paged), see Incidents.

What a Log Row Is #

A row is one result from one check. An uptime monitor checked from 12 probes writes 12 rows per round, a DNS monitor writes one row per sweep of its zone's nameservers, and a heartbeat monitor writes one row per ping received.

Rows are raw results. A single Down row does not mean an incident opened: outage confirmation, alert delay and grouping all sit between a failed check and a page.

Most rows carry 2 statuses, Up and Down. Two families add more:

  • DNS rows read amber Partial when the sweep finished but left a query unanswered or a health issue open, and amber Warning when the only records that differ from the baseline are Warning severity. Only a Critical change makes a DNS row Down.
  • Security rows sum up the whole scan in one word: red Infected when Google Web Risk flagged malware or phishing, an amber issue count such as 2 issues when other checks left issues open, amber Unknown when a check reached no verdict, and Clean, No issues or a grey Not checked otherwise.

The blacklist Check log on the monitor's own page also shows an amber Partial for a sweep that left a list unreadable. In this list a blacklist row is Up or Down.

Finding the Page #

Logs is a top-level sidebar item, not a child of Incidents, and the page's title is Logs.

Note: Logs and Audit Log are 2 different pages. Logs is monitoring events (probes, pings, sweeps), while People → Audit Log is the account-action trail: who invited a member, who deleted a monitor. See The Account Activity Log.

Every role except Accounting / Billing can open the page. A seat limited to monitor groups sees only those groups' checks, on the page and in every export. See Limiting a Member to Monitor Groups.

The Logs page: a one-line summary of checks, failures and monitors down now, the All checks and Failed only tabs, the search box and filter pills, and the table of individual checks grouped by day.Enlarge

The Summary Line #

The summary line above the table describes the view you are looking at right now, so changing a filter moves it: how many checks matched, how many of them failed and what share that is, and which monitors are down now.

The down-now count is live and ignores the date range. It counts only monitors that are checking: paused, suspended and in-maintenance monitors are excluded, not counted as down.

Filtering the List #

Two tabs, All checks and Failed only, sit at the left of the filter row. Beside them are a search box, 3 pill dropdowns and the date range. Each pill labels its own state: "All types", "3 types", "12 monitors", "All locations". A Clear all button appears once anything is narrowed. It leaves the tab as you picked it.

Search matches monitor names, not URLs or response bodies, and combines with the other filters.

Failed only lists failing checks. All checks lists passing and failing checks together. A DNS Warning row and a Security row with open issues count as passing, so they appear only under All checks, as does a Security row whose malware lookup failed (Unknown). A DNS Partial row counts as failed when a query went unanswered, and as passing when only a health issue is open.

Monitors lists your monitors grouped by type, alphabetically inside each group.

Locations lists probes grouped by region; expand a region to tick individual countries. The pill disappears when the only selected types are Server, Heartbeat and Security, because none of them records a probe location: Server and Heartbeat data comes in from your side, and a Security scan runs from Uptimia's servers. DNS and Blacklist keep the pill. Each sweep runs from one probe and records it, so you can filter their rows by that probe's country.

Date range offers Today, Yesterday, Last 7 Days, Last 30 Days, This Month, Last Month and a custom range. The default is Last 7 Days, remembered across pages for the rest of the browser session.

Which Types Produce Log Rows #

The type dropdown groups 11 families into 4 sections, plus one row on its own:

SectionTypes
Availability & performanceUptime, Speed, Transaction, API
Domain & certificatesSSL, Domain, DNS
ReputationSecurity, Blacklist
InfrastructureServer
(on its own)Heartbeat

All types toggles in one click, so the fast way to isolate a single family is to un-tick it and then tick the one you want.

The type filter open on the Logs page, scrolled to the top of its list: the All types row, Availability & performance with Uptime, Speed, Transaction and API, and Domain & certificates with SSL, Domain and DNS. Reputation, Infrastructure and Heartbeat sit further down the scrollable menu.Enlarge

What one row means differs per family, and so does the number in the Response column:

TypeOne row isThe drawer's headline reading
UptimeOne check from one probeResponse time
SpeedOne full page load from one probeFull page load
TransactionOne scripted browser runTotal run time
APIOne run of the request chainTotal run time
SSLOne certificate checkDays until expiry
DomainOne WHOIS lookupDays until expiry
SecurityOne security scanMalware and phishing: Clean, Infected or Unknown
DNSOne sweep of the zone's authoritative nameserversSlowest nameserver
BlacklistOne sweep of the blacklistsSweep time
ServerOne sample reported by the agentReported CPU
HeartbeatOne inbound pingRun duration, or lateness against the schedule

Real User monitors produce no log rows, so they have no chip in the filter and no per-check export. RUM collects timings from your visitors' browsers as a continuous stream rather than as discrete checks. Read it on the monitor's own page. See Real User Monitoring.

Heartbeat Pings Only Appear in a Heartbeat-Only View #

A heartbeat row has a different shape: a signal type and a source IP, not a probe status and a probe location. Uptimia returns heartbeat rows only when heartbeat is the whole type selection, and never mixes them into an all-types view.

The filter enforces that. Tick Heartbeat and every other type clears; tick another type while heartbeat is selected and heartbeat drops out. The page states the rule under the filter row: "Heartbeat pings are a separate stream — other monitor types are hidden."

A missing ping never surfaces in the default all-types view. To check whether a cron job reported in, select Heartbeat, or open the monitor's own Ping log card. See Heartbeat Monitoring.

Server Rows Have No Pass or Fail #

Server monitors report measurements: CPU, load, memory, disk. They do not report a pass or a fail. There is no pass/fail column behind a server row, so every server row renders as Up, and the Failed only tab drops server rows from the list entirely rather than listing them as failures. All checks keeps them.

The Failed only tab is never where you find a server problem. Filter to the Server type on All checks, read the metrics in the drawer, and take the verdict from the monitor's own thresholds and incidents. See Server Monitoring.

Reading the Table #

Rows are grouped by calendar day. Each day header carries its label: Today, Yesterday, or the date. When any of that day's rows failed, a red count follows, such as "· 3 failed", with "on this page" added when the page does not hold the whole range.

Columns are Time, Monitor, Result, Location and Response; the monitor name links to the monitor, and clicking anywhere else on the row opens the check detail drawer. Location shows a flag and city for probe-run checks, the source IP for a heartbeat ping, and — for server rows, which record no location. Response holds response time for probe checks and run time for heartbeats. Server and Security rows show —, because neither stores a timing. A server sample's CPU number is in the drawer.

An amber strip appears when 3 or more failing rows share a root domain and a probe country inside a 3-minute window. The rows have to be named as hostnames, such as "shop.example.com". The strip reads "4 checks failing on example.com from Germany · 14:31–14:33 — likely their side", and the band groups the rows without hiding any of them.

Monitors named in prose, such as "Payment Gateway", are never banded, because the banding reads the monitor name, not the URL.

Page size is 20, 50 or 100 rows.

The Check Detail Drawer #

Clicking a row slides out the check detail drawer. Its header carries the status pill, the monitor name, the target and that family's headline number. The target is the URL for uptime and speed monitors, and the certificate domain for SSL.

Every family shows the same first 3 facts: Monitor, Type and Time. A fourth names where the check ran: Location for the probe families, Ran from for API, Swept from for DNS and Blacklist, and Status for Server, which records no location. Each family then adds its own:

TypeWhat the drawer adds
UptimeStatus, Response time, URL and port, String send / String expect / Keyword expect where configured, Error, and a Response block with the raw response body
SpeedStatus, Response time, URL and port, and the same string and keyword facts
TransactionStatus, Total run time, Error
SSLValid or Invalid, Domain, Valid from, Valid to with days remaining, Issuer, Protocol, Error, and a Certificate lifetime meter scaled to the certificate's real lifetime
DomainWhether the lookup succeeded, Creation date, Updated date, Expiry date with days remaining
SecurityMalware and phishing (Clean, Infected, or Unknown when the lookup failed), Error, and a Checks section with every check of that scan: its result in that scan (such as No issues, an issue count with the worst severity, Unknown, Pending or Not run in this scan) and each issue still open after it, with its severity, evidence and how to fix it
HeartbeatSignal, Δ vs schedule (on time, or how many seconds late), Run time, Ping URL, User agent, Error
ServerCPU & load (CPU usage, 1m / 5m / 15m load), Memory (used, total, cached, swap), a Disk usage table per filesystem, plus process count and network RX / TX
APIResult (Passed or Failed), Run time, a Steps / Passed / Not run counter strip, and every step in chain order with its timing and error
DNSResult, Slowest server, counters for Queries / Servers / Answered / Changes / Health, and each nameserver's answers
BlacklistResult (Listed, Partial or Clean), Sweep time, counters for Queried / Listed / Info / Clean / Errors, and the informative list answers

On a Security row the drawer's status pill shows the same scan word as the table, while the headline reading is the malware and phishing verdict alone.

The check detail drawer open on a Security row of the Logs page: the 6 issues pill, the Clean malware and phishing reading, and the Checks section with a chip per check and the open findings, each with its severity, evidence and How to fix.Enlarge

Blacklist detail lists only the answers that say something, and counts clean lists in the strip without listing them.

The check detail drawer open on a failed uptime check, showing the Down pill, the monitor and URL, the response-time readout, the facts grid and the raw 503 response body for that single check.Enlarge

Opening the Drawer From a Monitor Page #

The same drawer opens from a monitor's own detail page. Scroll to the Check log card and click any row. The DNS page calls that card Sweep log, the heartbeat page calls it Ping log, and the drawer is identical, minus the Open monitor button.

The card shows the 10 most recent checks. The heartbeat Ping log shows the newest 20 pings. View all logs below it opens Logs pre-filtered to that monitor and its type.

The Check log card on a monitor's own page, listing its most recent individual checks with result, probe city and response time, and the View all logs link that opens Logs filtered to that monitor.Enlarge

Live Updates and Keyboard Triage #

The page refreshes every 60 seconds, as the line under the heading says, and rows that arrive on a refresh flash green.

The table supports keyboard triage when the drawer is closed and you are not typing in a field:

KeyAction
j / kMove focus down / up one row
EnterOpen the focused row's drawer
EscClose the drawer
/Jump to the search box

Exporting Checks #

The filtered list. Click Export CSV. The file is UptimiaLogs.csv, and it honors every active filter and the date range. It holds up to 100,000 rows, with the columns Name, Type, Location, Time, Status and Duration. A Security row's Status is the same scan word the page shows.

One check. From the drawer, open Export. Download PDF gives you that check's full detail printed to A4 as Monitoring-Log.pdf, and Download HTML gives you the same document as a standalone Monitoring-Log.html. Create Public Link publishes the check as a page on your Uptimia host, and anyone holding that link can open it without signing in.

The PDF is that same HTML with one difference. It opens every collapsible evidence block, so a printed page hides nothing behind a disclosure. All 11 log-producing families export this way, each carrying its own detail: API steps, DNS nameserver answers, blacklist list answers, server filesystems.

The Export menu inside the check detail drawer, offering the check as a PDF, as a standalone HTML file, or as a shareable public link.Enlarge

Warning: A public link needs no sign-in and cannot be revoked from the control panel. Share it as you would the check itself.

  • Server samples are kept for 24 hours. A server check older than a day cannot be exported, and its drawer says so: "This check's metrics are no longer available — agent samples are kept for 24 hours." Export a server check the day it happens, or read the numbers off the monitor page's charts.
  • A heartbeat ping URL is masked in every exported document, down to its last 4 characters: hb_····a1b2. A public-link export is served unauthenticated and the ping URL is the monitor's only credential, so the masked form shows enough to identify the monitor and not enough to fake a ping. The drawer itself, which is behind your login, shows the URL in full.

The format map for everything else Uptimia exports is in Exporting Your Data.

How Far Back the Logs Go #

Check history is readable for as long as your plan's retention window allows:

PlanReadable window
Free7 days
Trial30 days
Basic90 days
Professional365 days
Enterprise730 days

Upgrading widens the window instantly and retroactively. The older rows were never deleted, only out of reach, so the next page load returns them with no backfill and no waiting. Data and Log Retention has the full rules.

Two screens can disagree. The Uptime, Speed and Security charts on a monitor's own page silently raise a too-early start date to your plan's floor. The account-wide Logs page and its CSV do not, and neither do the other families' charts, so on a short-window plan a long custom range here can return rows an Uptime, Speed or Security monitor page will not show you. See Exporting Your Data.

Was this helpful?

Your answer helps the team improve this article.

Still need help?

Uptimia Bot answers right away and passes you to the team when it can't.

Source: https://help.uptimia.com/help/a/monitoring-logs