PI Nexus+ / Reference
Host Monitoring Reference
This reference lists the issues, limits, rights, ports and timings of host monitoring. The Administration Guide and the User Guide link here instead of repeating these values.
Overview
This reference lists the issues, limits, rights, ports and timings of host monitoring. The Administration Guide and the User Guide link here instead of repeating these values.
Host issues
Host issues appear in the Issues page list, marked Hosts (the Category filter has no Hosts entry), on the Issues tab of the host dialog, and on the sensor named in the last column. Values marked adjustable are set under Admin > Hosts > Limits or per host (see Default limits).
| Issue | Severity | Raised when | Resolves when | Sensor | Typical fix |
|---|---|---|---|---|---|
| Low disk space | Warning below 10 % free, Error below 5 % (adjustable; per disk also in GB) | A disk is below its limit on a poll | The disk is above the limit again | Disk and the drive, for example Disk C: | Free space or extend the disk; set a disk's own limit for large archive disks |
| Disk filling up | Warning | At the trend of the last 7 days, the disk is full within 30 days (adjustable, 0 = off). Needs at least 24 hours of data. Not raised while Low disk space is open | The forecast is beyond the limit | Disk and the drive | Find what is growing; plan more space |
| High CPU | Warning; optional Error. Off by default | The average CPU over the window (default 30 minutes) is above the limit; one missed poll is tolerated | The average is more than 5 points below the Warning limit | CPU | Find the busy process on the Processes section |
| High memory use | Warning; optional Error. Per host only, off by default | The average memory use over the window is above the limit | As for High CPU | Memory | Check the processes; do not switch this on for PI Data Archive or SQL Server hosts, which use memory as cache |
| Host not answering | Warning | Three polls in a row failed. Disk, CPU and memory issues keep their state meanwhile | A poll succeeds | Host | Check that the host runs and that WinRM answers; run Test access |
| Service not answering | Error | An availability check failed 3 times in a row (adjustable 1 to 10) | 2 good checks in a row (adjustable 1 to 10) | The check, for example PI Data Archive PISRV01 | Check the PI service on the host and the network path from the scanner server |
| Certificate expiring | Warning within 30 days; Error when expired or invalid | An HTTPS check's certificate | The certificate is renewed | The HTTPS check | Renew the certificate |
| Host restarted | Warning | The host's boot time moved by more than 2 minutes. Not raised inside a maintenance window | Automatically after 24 hours | Uptime | None needed if planned; otherwise check the Windows event log |
| Clock drift | Warning above 5 s, Error above 60 s (adjustable) | The host's clock differs from the scanner server's by more than the limit plus the measurement's uncertainty | The drift is within the limit | Clock | Fix the host's time synchronization |
| Process not running | Error | A watched process is missing in as many polls in a row as Not answering after failed checks (default 3) | The process runs again | Process · and the program name | Start the PI component; tick Forget the role processes and application pools seen so far after removing one on purpose |
| Windows service stopped | Error | A watched service with start mode Automatic is not running in 2 polls in a row. The message names the state and exit code | The service runs | The service name (group Services) | Start the service, or set an unused service to Manual or Disabled on the host |
| Windows service restarting | Warning | A watched service's process changed 3 times within 24 hours without a host restart | No further restarts in 24 hours | The service name | Check the service's own log and the Windows event log |
| Windows service disabled | Warning | A watched service was switched from Automatic to Disabled while monitored | Automatically after 24 hours | The service name | None needed if intended |
| Slow disk | Warning. Off by default | A disk's average access time (the slower of read and write) is above the limit over 30 minutes | The average is more than 5 ms below the limit | Disk latency | Check the storage and the load on the disk |
| Network errors | Warning. Off by default | A network interface reported errors in every poll of 30 minutes | A poll of the last 30 minutes had no errors | Network | Check the NIC, driver, cable and switch port |
| PI Backup failed | Error | The PI Backup Subsystem reports that its last backup failed | The next backup succeeds | PI Backup | Check the PI Data Archive message log and the backup target |
| PI event queue growing | Warning | The PI Snapshot event queue grew in every poll of 30 minutes | The queue stops growing | Event queue | Check the PI Archive Subsystem and disk performance |
| PI Buffer queue growing | Warning; Error above a size (adjustable, off by default) | The buffer queue of an interface node for one PI Data Archive grew in every poll of 30 minutes | The queue stops growing | Buffer · and the PI Data Archive | Check the connection from the node to the PI Data Archive |
| SQL backup old | Warning | A database has had no full backup for 2 days (adjustable, 0 = off). See the note below | A full backup is taken | Backup · PI System or Backup · other | Fix the SQL Server backup job |
| SQL log filling | Warning above 80 %, Error above 95 % used (adjustable) | A database's transaction log is above the limit | The log is below the limit | Log · and the database | Back up the log (FULL recovery) or check long-running transactions |
| SQL blocking | Warning. Off by default | A session has been blocked longer than the limit (5 minutes when switched on) | No session is blocked longer than the limit | Blocking | Find the blocking session in SQL Server |
| Application pool stopped | Error | An IIS application pool that was seen running is stopped | The pool runs | App pool · and the pool name | Start the pool in IIS Manager |
Note: SQL backup old covers online databases in FULL or SIMPLE recovery, except tempdb and snapshots. A database never backed up counts from its creation. Each SQL Server instance has at most two such issues: one for the PI System's databases (the AF database PIFD, and the PI Vision and PI Nexus+ databases PI Nexus+ connects to) and one for all others. Each names the three oldest.
Pausing or deleting a host resolves its issues without an email. A disk marked PI archive volume is named "PI archive volume" in its issues.
Monitoring board statuses
A host has the status of its worst sensor. The counts at the top of the board count sensors.
| Status | Meaning |
|---|---|
| Down | The host failed 3 polls in a row, an availability check reached its failed-checks limit, or a PI Interface or PI Adapter on the host reports Error |
| Error | An open Error issue on the sensor |
| Warning | An open Warning issue; a missed check or one or two missed polls; or an interface or adapter that reports Warning or Unknown |
| Stale | No new data for 3 intervals: 15 minutes for a polled host; for a host with availability checks only, 3 check intervals and at least 5 minutes |
| Up | Everything answers |
| Maintenance | The host is inside a maintenance window. Never counted as a problem |
| Paused | Monitoring is paused or access is not confirmed. Interfaces and adapters on the host are still monitored and can change the host's status |
Suppressed issues do not change a status, except that a host or check that does not answer is Down even when its issue is suppressed. Several status counts can be combined as a filter. Hosts in maintenance and paused hosts have no rows on the Alarms tab. On the Capacity tab, a disk that is not filling shows Not filling.
Host status on Servers > Hosts
| Status | Meaning |
|---|---|
| Monitored | The last poll succeeded |
| Not answering | The last poll failed |
| Paused | Monitoring is paused |
| No access | The access test has not passed |
| Waiting for first poll | Access passed; no poll yet |
Sensors per group
| Group | Sensors |
|---|---|
| Resources | CPU, Memory, Disk per drive, Clock, Uptime, Network, Disk latency, and Process · name while a watched process is missing |
| Availability | Host (polled hosts) and one sensor per availability check, for example PI Data Archive PISRV01 or TCP port 1433 |
| PI Data Archive | Event queue, PI Backup, Archive fill |
| PI Buffer | Buffer · and each PI Data Archive the node buffers to |
| SQL Server | Backup · PI System, Backup · other, Log · and each database, Blocking |
| PI Vision | App pool · and each IIS application pool |
| Services | Each watched Windows service |
| Interfaces | Each PI Interface with runtime monitoring on the host |
| Adapters | Each PI Adapter service on the host (the worst of its Collectors) |
A sensor the host does not deliver is left out. A host read from PI points has only CPU, memory and disk sensors, plus its availability checks.
Role readings
| Role | Read in each poll |
|---|---|
| PI Data Archive | Archived, snapshot and out-of-order events per second; PI Snapshot event queue; primary archive fill; PI Backup result (failed or not; the counters carry no backup time) |
| Interface node | Per PI Data Archive buffered to: queued events, events sent per second, connection; round trip from the node to up to 3 PI Data Archives (a ping run on the node) |
| SQL Server (counters) | Page life expectancy, batch requests per second, blocked processes, memory grants pending, per instance including named instances |
| SQL Server (T-SQL) | Per database: data and log size, growth, log used, age of the last full and log backup; blocked sessions. Only on PI Nexus+'s own SQL Server, the PI Vision SQL Server, and instances added under SQL Server instances |
| PI Vision | State and recycles of each IIS application pool; connections and requests per second per web site; w3wp memory |
Default limits
Set the defaults under Admin > Hosts > Limits. A host overrides Disk, CPU and Disk latency by clearing Use defaults in its Edit dialog; a disk overrides the host under Limits > Volumes. Disk settings apply in the order disk, host, default; Ignore this volume wins over all.
| Setting | Default | Range | Set in |
|---|---|---|---|
| Disk Warning below % free | 10 | 1 to 99 | Limits, host, disk |
| Disk Error below % free | 5 | 1 to 99, below Warning | Limits, host, disk |
| Disk Warning/Error below GB free | Not set | Above 0 to 100,000 GB | Disk only |
| Disk Forecast days (0 = off) | 30 | 0 to 365 | Limits, host, disk |
| Alert on sustained CPU | Off | Limits, host | |
| CPU Warning above % | 90 when switched on | 1 to 99 | Limits, host |
| CPU Error above % | Off | 1 to 99, above Warning | Limits, host |
| CPU Sustained minutes | 30 | 10 to 240, steps of 5 | Limits, host |
| Alert on sustained memory use | Off | Suggested 95 % over 30 minutes | Host only |
| Check every | 1 minute | 1 or 5 minutes | Limits |
| Not answering after failed checks | 3 | 1 to 10 | Limits |
| Recovered after good checks | 2 | 1 to 10 | Limits |
| Clock drift Warning above seconds | 5 | 1 to 3600 | Limits |
| Clock drift Error above seconds | 60 | 1 to 3600, above Warning | Limits |
| Alert on sustained disk latency | Off; 50 ms suggested | 1 to 10,000 ms | Limits, host |
| Alert on network errors | Off | Limits | |
| Error when a PI Buffer queue is above a size | Off; 100,000 events suggested | 1 to 1,000,000,000 | Limits |
| Full backup older than days (0 = off) | 2 | 0 to 365 | Limits |
| Log Warning above % used | 80 | 1 to 99 | Limits |
| Log Error above % used | 95 | 1 to 100, above Warning | Limits |
| Alert on blocked sessions | Off; 5 minutes when on | 1 to 1440 minutes | Limits |
These values are fixed: Host not answering after 3 failed polls; Certificate expiring 30 days ahead; Windows service stopped after 2 polls; Windows service restarting at 3 restarts in 24 hours; 30-minute windows for Slow disk, Network errors, PI event queue growing and PI Buffer queue growing. A changed limit applies from the next poll or check; past data is not re-evaluated.
Intervals and timings
| What | Interval or time |
|---|---|
| Resource poll (CPU, memory, disks, services, roles) | Every 5 minutes |
| Availability checks | Every minute (or every 5 minutes) |
| Access test after adding or editing a host | Within seconds; 3 hosts at a time |
| Disk forecast recalculation | Every hour |
| Discovery of role counters (PI, SQL Server, IIS) | At the access test and once a day |
| Monitoring page refresh | Every minute while visible |
| Red "data is old" banner on the Monitoring page | Data older than 15 minutes |
| No Windows service findings after a host starts | First 15 minutes |
| PI point value counted as missed | Older than 15 minutes or bad quality |
| Host restarted and Windows service disabled stay open | 24 hours |
| Issues opened during a maintenance window are emailed | Within a minute after the window ends |
| Scanner service watchdog email | After 10 minutes without a report from the Scanner service (checked every minute) |
| Reminder and escalation emails | Checked with every availability check cycle |
Timeouts
| Operation | Timeout |
|---|---|
| WinRM port test | 5 s |
| WinRM (CIM) operation | 20 s |
| One host's poll | 60 s |
| HTTP check, PI Data Archive or AF Server check | 15 s |
| TCP check | 5 s |
| SQL Server connect and command | 5 s; 10 s per instance in total |
Retention
| Data | Kept for |
|---|---|
| Five-minute samples and individual check results | 14 days |
| Hourly averages (with lowest and highest value) | 400 days |
| Restart markers | 400 days |
Ports
All connections start from the scanner server unless stated otherwise.
| To | Port | Used for |
|---|---|---|
| Monitored host | TCP 5985 | WinRM over HTTP (default) |
| Monitored host | TCP 5986 | WinRM over HTTPS (credential profile) |
| PI Data Archive | TCP 5450 | Availability check; PI points hosts |
| AF Server | TCP 5457 | Availability check |
| PI Vision | TCP 443 or 80 | Availability check of the PI Vision URL |
| SQL Server | TCP 1433 or the instance port | SQL Server reads over T-SQL (optional) |
| Any | As configured | Availability checks you add |
| From an interface node to its PI Data Archives | ICMP echo | PI Buffer round trip (shown only; Not available where blocked) |
Required Windows rights on each host
Grant these to the scanner service account, or to the account of the host's credential profile. No administrator rights are needed.
| Right | Without it |
|---|---|
| Member of Remote Management Users | WinRM refuses the sign-in |
Enable Account and Remote Enable on the WMI namespace Root\CIMV2 | Every reading is denied |
| Member of Performance Monitor Users | CPU, disk latency, network, process and role counters return no data; memory and disks still work |
| Read access to the Service Control Manager and to each PI-related service (optional; granted by the rights script) | Windows services are not read; PI processes are watched instead |
WinRM must be enabled on the host (winrm quickconfig).
Rights script parameters
Grant-PINexusHostAccess.ps1 runs in Windows PowerShell 5.1 (not PowerShell 7) as a local administrator.
| Parameter | Meaning |
|---|---|
-Account | Required. The account to grant, as DOMAIN\user, HOST\user or user@domain |
-Service | Further service names to grant read access to, comma-separated, for example 'MyOpcServer,OtherService' |
-SkipServices | Leaves out the Windows services step |
-WhatIf | Shows what would change without changing it |
The script finds the local groups by their well-known SIDs, so it works on any Windows language. It changes only what is missing and can be run again. It runs Enable-PSRemoting -SkipNetworkProfileCheck only when WinRM has no running listener. It prints the security descriptor of each service it changes, before and after; to undo one grant, run sc.exe sdset <service> "<before>".
The script refuses to run on a domain controller, which has no local groups: it stops before any step, also with -WhatIf, and changes nothing. Grant the rights on a domain controller through domain policy instead.
Group Policy
For many hosts, grant the same rights with a GPO linked to the servers' OU.
| What | Policy setting |
|---|---|
| Local groups | Computer Configuration > Preferences > Control Panel Settings > Local Users and Groups: one Group item for Remote Management Users (built-in) and one for Performance Monitor Users (built-in), action Update, Add the scanner service account. Restricted Groups also works, but replaces the whole membership |
| WinRM listener | Computer Configuration > Policies > Administrative Templates > Windows Components > Windows Remote Management (WinRM) > WinRM Service > Allow remote server management through WinRM: Enabled, IPv4 filter * or the scanner server's address |
| WinRM service | Computer Configuration > Policies > Windows Settings > Security Settings > System Services > Windows Remote Management (WS-Management): Automatic |
| Firewall | Windows Defender Firewall with Advanced Security, inbound rule Windows Remote Management (HTTP-In), TCP 5985, ideally limited to the scanner server |
| WMI namespace and service rights | No policy setting. Run Grant-PINexusHostAccess.ps1 -Account 'DOMAIN\svc-pinexus' as a computer startup script (Computer Configuration > Policies > Windows Settings > Scripts (Startup/Shutdown) > Startup, PowerShell Scripts tab, the account as script parameter), or as an immediate scheduled task running as SYSTEM (Preferences > Control Panel Settings > Scheduled Tasks). The script changes only what is missing, so it can run at every start |
After gpupdate /force (or the next policy refresh), restart hosts where WinRM was newly enabled, then choose Test access. Group membership applies to new sign-ins only.
Required SQL Server rights (optional)
The scanner service account reads SQL Server with Windows authentication. All reads are SELECT statements and DBCC SQLPERF(LOGSPACE); nothing is written.
| Right | Grant with | Without it |
|---|---|---|
| A login | CREATE LOGIN [DOMAIN\svc-pinexus] FROM WINDOWS; | Nothing is read over T-SQL; counters still are |
| VIEW SERVER STATE | GRANT VIEW SERVER STATE TO [DOMAIN\svc-pinexus]; | Log use and blocked sessions are not read |
| VIEW ANY DEFINITION | GRANT VIEW ANY DEFINITION TO [DOMAIN\svc-pinexus]; | File sizes and growth are not read |
SELECT on msdb.dbo.backupset | In msdb: CREATE USER [DOMAIN\svc-pinexus] FOR LOGIN [DOMAIN\svc-pinexus]; GRANT SELECT ON dbo.backupset TO [DOMAIN\svc-pinexus]; | Backup age is not read |
Access test steps
| Step | Required | Checks |
|---|---|---|
| Name lookup | Yes | The host name resolves from the scanner server |
| WinRM port | Yes | TCP 5985 or 5986 answers |
| WinRM sign-in | Yes | The account can sign in; names the account and transport |
| WMI access | Yes | Root\CIMV2 can be read |
| Memory, CPU counters, Disks | Yes | The core readings |
| PI points | Yes (PI points hosts only) | The mapped PI points can be read |
| Disk latency (optional), Network (optional), Processes (optional) | No | Performance counters |
| Windows services (optional) | No | How many services the account can read and how many are watched |
| PI Data Archive counters (optional), PI Buffer counters (optional), SQL Server counters (optional), SQL Server databases over T-SQL (optional), IIS counters (optional) | No | Role readings; listed only for hosts with that role |
An optional step that is not available never raises an issue; the host is monitored without it.
Access problems
Open the host's Access details (the Access label, or Access details in the row menu) and start with the first failed step.
| Step or symptom | Cause | Fix |
|---|---|---|
| Name lookup failed | The name does not resolve on the scanner server | Correct the address or the DNS entry |
| WinRM port failed | WinRM is off, or a firewall blocks TCP 5985 or 5986 | Run winrm quickconfig (HTTPS: create the HTTPS listener) and open the port from the scanner server |
| WinRM sign-in failed: logon rejected | Wrong account or password, or the account is disabled, locked or expired | Check the account; for a profile, enter the password again under Credentials |
| WinRM sign-in failed: access denied | The account is not in Remote Management Users | Run the rights script |
| WinRM sign-in failed: IP address | An IP address with the scanner service account or over HTTP | Add the host by DNS name, or use a credential profile over HTTPS |
| WinRM sign-in failed: certificate refused | The HTTPS certificate is not trusted or not issued to the host's name | Issue a certificate for the host's name from a trusted CA, or use Skip CA check / Skip name check |
| WinRM sign-in failed: password cannot be decrypted | Split installation without the shared secret certificate on both servers | Install the certificate on both servers, then enter the profile's password again |
| WMI access failed | No Enable Account / Remote Enable on Root\CIMV2 | Run the rights script |
| CPU counters failed, or an optional counter step not available | The account is not in Performance Monitor Users | Run the rights script; sign-ins made before the change do not see it, so test again |
| Windows services (optional) not available | Service rights not granted, or a PI component was installed after the script ran | Run the rights script again on the host |
| SQL Server databases over T-SQL (optional) not available | No login, or missing rights on the instance | Grant the rights in Required SQL Server rights (optional) |
| PI points failed | The PI Data Archive is not reachable, or the points are wrong | Check the PIPerfMon points and use Find points |
| Host shows Stale on the board | The Scanner service has stopped | Start the PI Nexus+ Scanner service |
| Disk issue on a disk that is fine with little free space | Percent limits do not suit a large disk | Set GB free limits for that disk |
| Windows service stopped for a service you do not use | The service is set to Automatic | Set it to Manual or Disabled on the host, or add it to Never watch |
The host's row menu also has Test access, Edit, Pause monitoring and Delete.
Watched Windows services
Watched services raise the three Windows service issues. Only a service with start mode Automatic is expected to run.
| Role or rule | Services |
|---|---|
| PI Data Archive | piarchss, pibackup, pibasess, pilicmgr, pimsgss, pinetmgr, pishutev, pisnapss, pisqlss, pitotal, piupdmgr |
| AF Server | AFService, PIAnalysisManager, PINotificationsService |
| SQL Server | MSSQLSERVER, MSSQL$instance, SQLSERVERAGENT, SQLAgent$instance |
| PI Vision | W3SVC, WAS |
| Interface node | pibufss |
| Any host | Every service whose program is installed under \PI\bin\, \PI\adm\, \PIPC\, \OSIsoft\ or \AVEVA\ (each PI interface and PI Adapter instance) |
| Added by you | Up to 25 names under Windows services > Also watch; Never watch excludes a service and wins |
pishutev stopping cleanly after startup is expected. A service that is uninstalled is not reported as stopped. A process whose service is watched is reported as that service, not as Process not running.
Watched processes
Processes are watched where Windows services cannot be read. A role process is expected only after it has been seen running on that host.
| Role | Processes |
|---|---|
| PI Data Archive | piarchss, pisnapss, pibasess, pinetmgr, piupdmgr, pimsgss, pibackup |
| AF Server | AFService, PIAnalysisManager, PIAnalysisProcessor, PINotificationsService |
| SQL Server | sqlservr |
| PI Vision | w3wp |
| Interface node | pibufss |
| Adapter node | OSIsoft.Data.System.Host |
| Added by you | Up to 25 program names (without .exe) under Processes > Also watch; always expected |
PI points read for a PI points host
Find points searches the chosen PI Data Archive for points whose instrument tag (or extended descriptor) is one of these counter paths with one of the host's names.
| Metric | Counter path |
|---|---|
| CPU | \\HOST\Processor(_Total)\% Processor Time |
| Memory | \\HOST\Memory\Available MBytes or \\HOST\Memory\% Committed Bytes In Use |
| Disk free %, per drive | \\HOST\LogicalDisk(D:)\% Free Space |
| Disk free MB, per drive | \\HOST\LogicalDisk(D:)\Free Megabytes |
Memory shows as used percent only when the host's memory size is known from an earlier WinRM reading. A disk with only % Free Space has no size, no GB limit and no forecast. A host has at most 60 point mappings.
Availability checks
| Check | Created | How it tests |
|---|---|---|
| PI Data Archive name | Automatically for a PI Data Archive target | TCP 5450, then a server-time request. TCP only for OpenID Connect targets |
| AF Server name | Automatically for an AF Server target | TCP 5457, then a server-time request. TCP only for OpenID Connect targets |
| PI Vision name | Automatically for a PI Vision instance | Opens the URL with Windows authentication; at most 3 redirects on the same host |
| HTTP address | By you | Opens the address. Sends Windows credentials only when the address names the host, or with Send Windows credentials |
| TCP port | By you | Connects to the port (1 to 65535) |
Checks run from the scanner server, also while the host's WinRM access fails. In the Availability checks dialog, Switch off stops a check and Delete removes one you added.
Notifications
| Setting | Where | Default |
|---|---|---|
| Host monitoring delivery rule | Admin > Notifications > Delivery Rules | Off |
| New issues, Severity escalations, Recoveries after an alert | Same rule | On once the rule is on |
| Minimum severity | Same rule | Warnings and errors (or Errors only) |
| Reminder | Same rule | Off; every 1, 2, 4, 8, 12 or 24 hours |
| Escalation | Same rule | Off; after 15 or 30 minutes, or 1, 2 or 4 hours |
| Host monitoring email type | Each recipient | Ticked |
| Host monitoring escalations email type | Each recipient | Not ticked |
Reminders and escalations apply only to open Error issues that are not acknowledged or suppressed and not inside a maintenance window. A reminder repeats after the chosen time since the last email about the issue; an escalation is sent once per occurrence. Maintenance windows suppress all host emails. The scanner service watchdog sends PI Nexus+ scanner service is not running and running again to the Host monitoring recipients when the rule is on.
Limits on size
| Item | Limit |
|---|---|
| Hosts added in one Add N selected | 100 |
| Hosts polled in parallel | 8 |
| Availability checks run in parallel | 32 |
| Processes or services added per host (Also watch, Never watch) | 25 each |
| SQL Server instances added per host | 10 |
| Databases read per SQL Server instance | 200 |
| PI point mappings per host | 60 |
| Address of an HTTP check | 512 characters |
| Maintenance window | 5 minutes to 31 days; weekly windows up to 7 days |
| Maintenance note | 512 characters |
| Rows on the Alarms tab | 500 (Download CSV includes all) |
Uptime figures
| Figure | Meaning |
|---|---|
| Availability | Share of good checks in the period, each hour weighted by how often it was checked. Hours that overlap a maintenance window are left out. Green at 100 %, amber from 99 %, red below |
| Average response | Average time of the answered checks |
| Outages | Runs of failed checks as long as it takes to raise Service not answering (3 by default). A host's outages are its checks' outages merged |
| Longest outage | Longest such run |
Who can do what
| Action | Role |
|---|---|
| View Monitoring and Servers > Hosts | Viewer |
| Acknowledge or suppress host issues; Download CSV | Operator |
| Everything under Admin > Hosts; host email settings | Admin |
Support bundle
The support bundle includes host-monitoring.json: hosts, access test results, last poll state, connection settings, checks, limits and credential profiles by name and account. Passwords, and user names or passwords in check addresses, are never included.
