Troubleshooting
Troubleshooting
Find your symptom, then follow its entry. Each entry has the likely causes, the steps, and how to confirm the fix. If you are stuck, collect diagnostics (Collecting diagnostics) and write to support@yuvired.com.
Two agent names. The server's own agent is the Windows service YuviRedAgent. On every other
computer the agent's service is SyntinelRedAgent (shown in Services as YuviRed Endpoint Agent),
and on every computer, the server included, the agent writes its log under C:\ProgramData\SyntinelRed
(the folder name is the product's earlier name; it will change in a later release).
Commands run in Windows PowerShell as administrator on the server. The server command is long because nothing is on the PATH; it is written out in full where needed:
& "C:\Program Files\YuviRed\runtime\python.exe" "C:\Program Files\YuviRed\runtime\yuvired_service.py" <command>
Contents
- The server: service will not start, console unreachable, certificate warnings, port in use, database failed, disk full, time skew, sign-in problems
- Licensing: code refused, activation service unreachable, locked console, server changed
- Computers (endpoints): not appearing, stale, log codes, token expired, rotation failure
- The browser extension
- Notifications
- Reports
- Where the logs are
1. The server
1.1 The YuviRed service does not start (or stops at once)
Symptom. In Services, YuviRed is Stopped, or starts and stops again. The console does not open.
The installer, if just run, said the install failed.
Likely causes.
- The database service
YuviRedPostgresis not running.YuviReddepends on it. - The boot stopped at a step. The boot status file names it.
- The data folder is not safe: it, or a folder in it, was replaced by a link or is not owned by an administrator, or its permissions were changed.
- The database is newer than the program (after installing an older build over a newer one, or after an update that was rolled back once the new version had already changed the database).
- The identity key does not open: the data folder was copied from another computer or Windows
installation. The boot status names the step
provision_identityand "the appliance private key could not be loaded" (Administration, section 2.2). - Port 8443 is taken (see 1.4) or the disk is full (see 1.6).
- The hardware identity changed (see 2.4). The console is locked, but the service itself still runs.
Steps.
Read the boot status:
Get-Content C:\ProgramData\YuviRed\status\boot-status.jsonYou should see:
"phase": "YuviRed Ready"when healthy. Otherwise the file shows the phase it stopped in (Preparing,Secrets Ready,Database Ready,Schema Current,Platform Bound,Identity Created,Registration Pending,Restricted,Clone Suspected (Restricted),Error) and, forError, which step failed. The file holds no secrets.Start the database first, wait ten seconds, then start the server:
net start YuviRedPostgres net start YuviRed- Open Event Viewer, Windows Logs, Application, and look at sources
YuviRedandYuviRedAgent. An error "YuviRed stopped at phase ..." names where it stopped. - Read the end of
C:\ProgramData\YuviRed\logs\syntinel-red.logand the newest file inC:\ProgramData\YuviRed\pg\data\log. - If the log says "Can't locate revision", the installed program is older than the data. Install the newer version over it.
- If the boot status or the log says a path "is a link" or "was not created by an administrator", something replaced a YuviRed folder. Do not delete it. Collect diagnostics and write to support.
- Run the support collector before you restart the computer or reinstall.
How to confirm. boot-status.json reads YuviRed Ready, all three services read Running after a
minute, and https://127.0.0.1:8443 opens.
Common problems. If the service starts but stops after a few seconds, read steps 3 and 4. Do not reinstall as a first step: a reinstall over a damaged install can leave a partial state.
1.2 The console does not open
Symptom. The browser says "This site can't be reached", "connection refused" or times out.
Likely causes and steps.
| Cause | Check and fix |
|---|---|
| The service is not running | See 1.1. |
| You are on another computer, and the server is in its default mode | By default the server answers only on the computer itself (https://127.0.0.1:8443). To serve other computers, turn on network mode on the server: ... yuvired_service.py network --name <dns name or fixed IP>. See the Installation guide. |
| Network mode is on but the firewall blocks | The rule YuviRed Server (TCP 8443) applies to the Domain and Private network profiles only. A network set to Public is not allowed. Run ... network --status to see the profile and whether the rule is present. Set the connection's network profile to Private, or add a rule of your own. |
| The name does not resolve | From the other computer, Resolve-DnsName yuvired.northwind.example. Fix DNS, or use the address 192.0.2.10. |
| A firewall or proxy between | Test with Test-NetConnection 192.0.2.10 -Port 8443 from the other computer. |
| Wrong address | Use https://, and port :8443. |
How to confirm. The sign-in page opens from the other computer. Test-NetConnection ... -Port 8443 shows
TcpTestSucceeded : True.
1.3 Browser certificate warnings
Symptom. "Your connection is not private", "NET::ERR_CERT_AUTHORITY_INVALID", or a name mismatch.
Likely causes.
- Expected on first use. The server makes its own certificate and no browser knows it. It is valid for
825 days and names the computer's name,
localhost,127.0.0.1and the names and addresses you gave to network mode. - Name mismatch. You opened the server by an address or name that is not in the certificate.
- Wrong clock on the server or on your computer (see 1.7): the certificate looks not yet valid or expired.
Steps.
- For the first sign-in, proceed past the warning (Advanced, then continue). It is your own server.
- For a name mismatch, open the server by a name or address in the certificate, or run
... network --name <the name you use>so the certificate includes it. **A new certificate stops enrolled computers trusting the server** (they trust exactly the old one): they need a new Add Computer download. Use a DNS name from the start to avoid this. - To remove the warning, import the server's certificate (
C:\ProgramData\YuviRed\config\tls-cert.pem, the public part only) into the trusted store of the computers' browsers, by your own means. Never copytls-key.pem.
How to confirm. The address bar shows the certificate without a warning, or you chose to proceed once.
1.4 Port 8443 or 5440 is already in use
Symptom. The service stops at start; the log says it could not bind; or the installer's wait fails.
Steps.
Find the program that holds the port:
Get-NetTCPConnection -LocalPort 8443 -State Listen | Select LocalAddress, OwningProcess Get-Process -Id <OwningProcess>Use
5440for the database port.- If it is another program, stop it or move it to another port. YuviRed's ports are fixed. A way to change them is not documented (not verified).
- If it is a leftover
pythonserviceorpostgresof YuviRed, stop theYuviRedandYuviRedPostgresservices, wait, and start them again.
How to confirm. Get-NetTCPConnection -LocalPort 8443 -State Listen shows the YuviRed service's
process (pythonservice).
1.5 The database service failed to start
Symptom. YuviRedPostgres will not start, or the console shows server errors; the YuviRed log
says it cannot connect to the database.
Likely causes. Disk full. The data folder or its permissions were changed. A restore that left
config\pg.pwfile and pg from different times. Antivirus holding the files. A crash that left the
database needing recovery (it usually recovers itself at start).
Steps.
- Read the newest file in
C:\ProgramData\YuviRed\pg\data\log(names arepostgresql-<date>.log). The last lines say why. - Check free space (
Get-PSDrive C). If it is nearly zero, see 1.6. - Check that nothing else (antivirus, backup software) is locking the folder: exclude
C:\ProgramData\YuviRedfrom scanning and from file-level backup while the service runs. - Start it again and wait a minute: after a crash it replays its log, which can take time.
- Do not delete files in
pg\data, do not run a repair tool on it by hand. Collect diagnostics, then write to support.
How to confirm. YuviRedPostgres is Running, port 5440 listens, and YuviRed starts.
1.6 The disk is full
Symptom. Services stop; the database log says "No space left on device"; the console is slow or errors.
What fills the disk. The database (raw telemetry is kept 30 days by default, and the agent reports every web connection of every program), the logs (bounded), report files (capped at 2 GB), and your own copies of the data folder.
Steps.
- Free space outside YuviRed first: other files, old copies, installers.
- When there is room to start, start
YuviRedPostgresthenYuviRed. - In the console, Configuration, Data & Retention: lower Raw telemetry (for example from 30 to 14 days), then, under Telemetry Retention, click Run now (dry run) and then Run now. A database does not shrink the disk file at once: space is reused, and may only be returned later.
- Move report files off with Reports, Delete files.
- Add disk space for the long term.
How to confirm. Free space is above 15 percent, no new "No space" lines appear in the logs, and the console loads.
1.7 Time skew
Symptom. MFA codes are refused although typed correctly; the certificate looks "not yet valid"; a paid license seems to end early or late; sign-ins end sooner than expected.
Cause. The server's clock (or a phone's, or a computer's) is wrong. The server trusts its own clock: moving it back extends a paid license's end date, moving it forward ends it early. Authenticator codes change every 30 seconds.
Steps.
w32tm /query /statusshows the last sync. Repair:w32tm /resync. In a domain the server gets time from the domain; elsewhere, set an internet time source in Windows Date & time.- On a virtual machine, turn on the host's time sync, or use Windows time, not both fighting.
- Fix the phone's clock to automatic.
How to confirm. MFA codes are accepted, and w32tm /query /status shows a recent sync.
1.8 Sign-in problems
| Symptom | Cause and fix |
|---|---|
| "Invalid credentials" | Wrong password. A temporary password given by an administrator stops working after 72 hours and the console then also says "Invalid credentials". Ask an administrator to reset it again. |
| "Too many failed login attempts. Please wait and try again." | 8 wrong attempts for the same address and account within 5 minutes. Wait 5 minutes. Restarting the YuviRed service clears the counters. |
| "Session Expired" box | Your session ended (time limit, or an administrator reset your password or ended your sessions). Click Log In Again. |
| The sign-in code is refused | Check the phone's time (1.7). Use a recovery code, or ask another administrator to reset your MFA. |
| Nobody can sign in | See Administration, section 5.3, "When the only administrator is locked out". |
2. Licensing
The license screen opens when the server is not activated after its 7 days, or when the license is not in force. Enrolled computers keep reporting and the extension only monitors, so staff are never blocked because of a license. New computers and new AI agents are refused until a license is active.
2.1 An activation code is refused
Symptom. After Activate, the screen shows a message, then "Need help? Email support@yuvired.com and include the installation ID."
Every refusal the service can give, with its code (the code is what support sees):
| The screen says | Code | Meaning and what to do |
|---|---|---|
| "That activation code is not valid, was already used on another server, or has expired." | invalid_code | A code works once, for one server, for 7 days from when it was issued. Check you typed it exactly; get a new one at https://yuvired.com/download. |
| "Your company already has 2 YuviRed Lite servers. To add another server, contact YuviRed support and tell us why you need it." | server_allowance_used | The company limit is reached (counted by installation ID, over time). Use the contact YuviRed link, or write to support@yuvired.com. If you are replacing a server, say so and give both installation IDs. |
| "Too many activation attempts. Try again in an hour." | rate_limited | Wait an hour. |
| "Activation is temporarily unavailable. Try again later, or use a license file." | signer_unavailable | YuviRed's side. Try later, or use the license file route (2.2). The code is not used up. |
| "The server sent no valid installation ID." | invalid_installation | Restart the YuviRed service and open the license screen again. If it persists, write to support. |
| "The request is too large..." / "The request must be a JSON object." | too_large, bad_request | A fault on the way (a proxy altering the request). Try the license file route. |
| "Only a YuviRed administrator can activate this server. Sign in with the administrator account created when YuviRed was installed." | (403 from your server) | Only the first account (super admin) can activate. Sign in with it. |
| "That file is too large to be a YuviRed license file." | n/a | You chose the wrong file. |
| "The license was stored, but the console is still locked. Email support@yuvired.com." | n/a | Write to support with the installation ID. |
The website (https://yuvired.com/download) can give these answers when you ask for a code: "Use your work
email address. Free email addresses cannot download YuviRed Lite..." (free_email_domain), "Two people at your
company already use YuviRed Lite..." (email_allowance_used), "Your company has reached today's limit..."
(daily_limit_reached), "We cannot accept this email address..." (domain_not_allowed), "That code is not right,
or it has expired. Ask for a new one." (the 6-digit email code, valid 30 minutes, 5 tries). The limits are 2 servers
and 2 verified people per company, and 3 codes per day.
Confirm the fix. The screen says License activated, and Configuration, License shows YuviRed Lite, Active.
2.2 The activation service cannot be reached
Symptom. "YuviRed could not be reached. Check this server's internet access and try again, or use a license file." or "Online activation is not set up on this server. Use a license file instead."
Likely causes. The server cannot reach https://activate.yuvired.com on TCP port 443; a firewall or
proxy; a proxy that replaces certificates; no internet; DNS.
Steps.
- From the server:
Test-NetConnection activate.yuvired.com -Port 443. Fix what blocks it. - Open
https://activate.yuvired.com/v1/healthin a browser on the server. A healthy service answers. - If the server must stay offline, use a license file: on any computer with internet, open https://activate.yuvired.com/license-file, enter the activation code and the installation ID, download the file, then on the server choose Upload license file.
Confirm the fix. License activated. After activation the server checks in once a day. If it cannot, it keeps working, and Last check-in stays at its last value.
2.3 The console is locked
The title tells you why.
| Title | Meaning | What to do |
|---|---|---|
| Activate YuviRed | The 7 days to activate ended. | Activate (code or file). A server that was activated once never gets the 7 days again. |
| Your YuviRed license has expired (Pro, Enterprise) | The end date passed. No grace. | Renew at https://yuvired.com/request or write to support. Activate the new code or file. |
| This YuviRed license was revoked | YuviRed revoked it ("YuviRed revoked the license for this server...") or an administrator did ("...was revoked on this server..."). | Write to support with the installation ID. A new code (a new serial) unlocks it. A revoked license cannot be used again. |
| This YuviRed license could not be verified | The stored license no longer verifies. | Activate a new code or file; if that fails, write to support. |
| This server needs a new license | See 2.4. | See 2.4. |
"YuviRed could not finish checking this server's hardware. Restart the YuviRed service, then try again."
means exactly that: restart the YuviRed service, then open the license screen again before you ask for
anything.
2.4 "This server needs a new license" (server_changed)
Symptom. After a restore, a copy, a move, or a change of the virtual hardware, the console shows "This server's hardware changed, so YuviRed treats it as a copy of another server."
Cause. The server records the computer's hardware identity (firmware serial numbers) next to its
installation ID and checks it at every start. If it changes (a virtual machine restored as a new
machine, imported or copied, a motherboard change, some hypervisor upgrades), a server that held a
license gives itself a new installation ID and locks (the audit log records license.server_changed
with the old and new IDs). The license is bound to the old one. YuviRed cannot tell a legitimate move
from a copy. Moving back does not undo it: it counts as another change.
Steps.
- If this is a copy of another YuviRed server, shut the copy down.
- If you moved or restored this server on purpose, copy the installation ID from the screen (Copy) and write to support@yuvired.com: say what you did, and give the old installation ID if you have it.
- Activate the new code or license file support sends you, as in 2.1.
- Meanwhile your computers keep reporting.
Confirm the fix. The screen reads License activated and the console opens.
3. Computers (endpoints)
3.1 A computer does not appear in Endpoints
Likely causes and steps.
| Cause | Check and fix |
|---|---|
| The Add Computer download expired | Its token lasts 24 hours. The agent log says so ("The Add Computer download has expired. Make a new download on the server..."). Make a new download. |
| All seats are used | The agent log says "All computer seats on the server are in use. Uninstall the agent from a computer you no longer use and remove that computer under Endpoints on the server, then restart this agent. If the Add Computer download is more than 24 hours old, make a new one instead." Lite allows 10 computers plus the server. At 10 of 10 the console refuses a new Add Computer download ("Only 0 computer seats are free."), so free a seat first. |
| It waits for an administrator | "The server is waiting for an administrator to confirm it under Agent Deployment, Add Computer. Restart this agent after that." Confirm it there, or under Adoption Requests, then restart the agent. |
| The computer cannot reach the server | From it: Test-NetConnection 192.0.2.10 -Port 8443. Check the address in the download, DNS, the server's network mode and firewall (1.2). |
| The computer does not trust the certificate | Happens after the server's certificate changed (1.3). Make a new Add Computer download. |
| The license is not in force | New computers are refused while the license is expired, revoked, not valid or the server not activated. See 2. |
| The agent service is not running | On the computer, Services, YuviRed Endpoint Agent (service SyntinelRedAgent), or read the agent log. |
How to confirm. The computer appears under Endpoints with Online within a few minutes, and the agent log shows a successful check-in.
3.2 A computer shows Stale or Offline
Status comes from the last check-in: Online up to 3 minutes, Stale up to 10 minutes, Offline beyond that (or never seen). The agent checks in every 60 seconds.
Steps.
- Is the computer on, awake and on the network? A laptop that sleeps is offline.
- Is the agent service running on it (
SyntinelRedAgent, shown as YuviRed Endpoint Agent; on the server itselfYuviRedAgent)? Start it. - Read its log (
C:\ProgramData\SyntinelRed\logs\syntinel-agent.log): "Cannot reach backend at https://... (...)" means network or name, and the bracket gives the reason; a 401 or 403 means its credential is refused (3.3). - Is the server healthy (1.1)? If many computers are offline at once, look at the server first.
- After repeated failures the agent slows down to one try every 5 minutes at most. Restart the service to force a try now.
How to confirm. The row turns Online within a few minutes.
3.3 What the numbers in the agent log mean (401, 402, 403, 410)
| Code | Where you see it | Meaning | What to do |
|---|---|---|---|
| 401 | Agent log, or in a script or browser console | The server does not accept the credential presented, or no credential was sent. | The agent first tries every credential it holds. If it keeps failing: the computer was removed under Endpoints, its credential was rotated or replaced, or the server's database was restored from before. Add the computer again (3.5). |
| 403 | Agent log at registration or check-in | The server refuses this computer. The agent log gives the reason in its own words and a fixed hint. Reasons: license seat limit reached, token expired, enrollment cap reached, invalid token, server identity changed, a decommissioned endpoint, a license revoked, expired or not valid. | Follow the hint in the log. See 3.1. |
| 402 | The console, a script calling the API | The console is locked because the license is not in force. This is a console answer. Computers keep reporting; they are not stopped by it. | Fix the license (2.3). If you see 402 in an agent log, that is not an expected message: collect diagnostics (not verified further). |
| 410 | Agent log during enrollment, or at check-in | During enrollment, "Campaign no longer exists": the Add Computer download (a campaign) that this computer was installed with has been completed, revoked or deleted. At check-in, the computer was retired (decommissioned) on the server. | Make a new Add Computer download and run its installer; for a retired computer, reactivate it under Endpoints first. |
| 426 | Agent log | The agent is older than the server's minimum agent version, or the server's connection security does not allow its connection. | Update the agent with a new Add Computer download. |
3.4 A token or download has expired
An Add Computer download carries a token that lasts 24 hours and can enroll a limited number of computers. When it has expired or its limit is reached, the agent log says so. Make a new download: Agent Deployment, Add Computer, then run the new installer on the computer. A computer that is already enrolled keeps its identity and uses no second seat. Delete old copies of the zip.
3.5 Credential rotation failed
Where. Agent Deployment, Endpoint Credential Rotation. Each computer has a rotation state:
| State | Meaning | Action |
|---|---|---|
| Pending | Waiting: the computer is offline, below agent 1.7.20, or has no credential of its own yet | Wait; agent 1.8.2 supports rotation. |
| In Progress | The new credential was delivered; waiting for the next check-in | Wait a few minutes. |
| Completed | Verified: the new credential works, the previous one is refused | None. If the row adds "awaiting first check-in" and ten minutes pass, check that computer. |
| Failed | A terminal problem | Click Retry (or Retry all failed/expired). |
| Expired | The window closed before the computer came back | Retry, with a fresh window. |
| Skipped | Already in another rotation | None. |
Steps.
- Open View on the operation and read the reason for the failed computer.
- If it is offline, bring it back, then Retry.
- If the computer cannot authenticate at all (the agent keeps its previous credential for 30 days after a completed rotation; beyond that, or after a snapshot revert or restored database, nothing can reach it): uninstall the agent from the computer and add it again with a new Add Computer download.
- A computer with the extension may show a short gap in prompt monitoring after a rotation. It recovers within a few minutes by itself.
How to confirm. The row reads Completed with "new credential verified, previous credential rejected", and the computer stays Online.
4. The browser extension
The extension (AI Prompt Monitor 0.5.5) works in Chrome and Edge. Firefox is not supported. Open its Options page: it shows a Connection card with the headline, the Product, Server, Transport, Policy, Ruleset and Last sync. Copy status for support at the bottom copies a text with versions and counts only (no prompts, findings or passwords).
4.1 The extension is not installed on a computer
Causes. In this release nothing installs the extension into Chrome or Edge by itself. The console setting Remote Browser Extension Wiring is off by default and, on a YuviRed Lite server, has no effect: the server does not serve a signed extension package, so it answers "disabled" to every computer. Store installs are planned and are not available until the store listings exist. Firefox is not supported.
Steps.
- On a computer added with Add Computer, a person with local administrator rights loads the extension
from
C:\Program Files\YuviRed Agent\extensionwith Load unpacked (chrome://extensionsoredge://extensions, Developer mode on). Use that folder, not a copy: a copy never updates. - Each person ticks the consent box on the options page, unless your administrator records it by browser policy.
Confirm. The extension appears in the browser's list and the options page shows a Connection card.
4.2 "No server" or "Not connected"
| Headline or row | Meaning | Fix |
|---|---|---|
| Not connected. Consent and save to start | Consent is not given or not saved | Tick "I consent to local inspection of prompts and attachments on the monitored sites above", Save. |
| Waiting for the first policy sync | Consent given; no policy pulled yet | Wait a minute; Save again. |
| Not synchronized, Last sync "failed: ..." | The last pull failed; the last known good policy stays active | Read the reason below. |
| Server: not set | No server address and no local agent found | Install the agent on the computer, or enter the server's HTTPS address in the options. |
Reasons shown after "failed:": it could not reach the server (check the network, and that the YuviRed agent is still installed, because on Lite the agent carries the browser's requests; or, if Transport says direct HTTPS, that this browser trusts the server's certificate); the server rejected the request (the computer is no longer enrolled, or its credential changed); the server returned an error; or the answer could not be read.
Steps. Confirm the agent service (SyntinelRedAgent) runs on that computer and the computer is
Online in the console (3.2). If the computer was removed from Endpoints or its credential rotated, add it again
(3.5). Then Save on the options page.
Confirm. The headline reads "Connected to YuviRed Lite" and Last sync reads "OK".
4.3 Policy: blocked, or nothing happens
- A person was blocked or warned. Policy is set in the console under Configuration, **Prompt Protection. A stronger policy than **Monitor Only warns or blocks. Read the policy and, if needed, change it there. A first install starts in Monitor Only, which shows nothing on the page.
- Nothing is blocked but you expected it. The extension only monitors when the license is not in force (expired, revoked, not valid, or the server not activated); no block is applied then. Fix the license (2.3). The options page's Policy row shows what the browser applies. "Company policy not received: the answer your administrator set for this case applies" means the policy could not be fetched, and the "if the policy cannot be fetched" answer is used.
- A page's first send is held a moment. Until a page learns its policy (up to 5 seconds), a send is held. This is expected.
Confirm. The Policy row on the options page shows the preset and revision you set.
5. Notifications
Notifications go out for new findings at or above Notify me about, subject to the quiet window and the hourly cap. The Test buttons ignore those brakes.
Where. Configuration, the Notifications chip, and the panel Notifications, Test Send below.
| Symptom | Likely causes | Steps |
|---|---|---|
| Test says "Channel 'email' is not configured yet." | Host and recipient (email) or the URL (Slack, Teams, webhook) are empty or unsaved | Fill them in and click Save Changes first. |
| Email test says "failed: ..." | Wrong mail server, port, login or TLS setting; the server cannot reach the mail server; the mail server refuses | The reason is the mail library's own text. Check Mail server, Mail server port (587 is preset), Mail username, Mail password, Encrypt mail connection (TLS), Send from address, Send alerts to. From the server: Test-NetConnection mail.northwind.example -Port 587. Check the mail server allows relay from this server. |
| Slack or Teams test fails | The webhook address was revoked, mistyped or pasted with spaces; outbound blocked | Create a new incoming webhook in Slack or Teams and paste it into Slack or Teams webhook URL. Check the server can reach the internet on 443 (or goes through your proxy). |
| Webhook test fails | The receiving system refuses, a certificate it does not trust, wrong address | Check the URL and the receiver's log. |
| Test works but real alerts never arrive | Notify me about is above the alert's severity; the quiet window; the hourly cap (Max notifications per hour); Don't repeat the same alert for hides repeats | Lower Notify me about, raise the cap, and check that the alert is new. |
| It worked and stopped | A password changed, a webhook was rotated, the mail server moved | Re-run Test. |
How to confirm. The channel's tag reads configured, Test says sent, and the message arrives
("... this is a test notification from the Configuration page...").
6. Reports
| Symptom | Likely causes | Steps |
|---|---|---|
| The report is empty or a section says "Not included: your role cannot view this" | You lack the permission for that area; or there was no activity in the period; or retention removed the data (raw telemetry is kept 30 days by default) | Ask an administrator to generate it, check the period, and check the retention settings (Administration, section 7). |
| Numbers look low | Computers were offline or not enrolled; the agent has no offline queue (a failed check-in is lost) | Check Endpoints for the period. |
| It takes a long time | A long period (the longest allowed is 400 days) and a large database; the server runs a limited number of reports at once (the exact limits were not verified here) | Choose a shorter period; wait for another report to finish. |
| "You already have reports being generated. Wait for one to finish." | Your own queue is full | Wait. |
| Status failed | The service restarted while it ran; or an internal error (a short reference is shown) | Generate again. If an internal error repeats, send the reference to support. |
| Download says the files were deleted | Report files are kept 30 days by default and total 2 GB | Re-run as new snapshot. |
Confirm. The report's status is complete and the file downloads.
7. Where the logs are
| What | Where |
|---|---|
| Server log | C:\ProgramData\YuviRed\logs\syntinel-red.log (10 MB, 5 old files) |
| Server agent start logs | C:\ProgramData\YuviRed\logs\agent-start.log, local-enroll.log |
| Database log | C:\ProgramData\YuviRed\pg\data\log\postgresql-<date>.log |
| Boot status | C:\ProgramData\YuviRed\status\boot-status.json |
| Agent log (every computer, including the server's own agent) | C:\ProgramData\SyntinelRed\logs\syntinel-agent.log (5 files of 5 MB; the earlier name, see the top of this page) |
| Installer logs | %TEMP% and C:\Windows\Temp, names start YuviRed_Lite_ |
| In the console | Admin, System Logs |
| Windows Event Log on the server | Event Viewer, Windows Logs, Application, sources YuviRed and YuviRedAgent (the agent's own messages may also appear under SyntinelRedAgent) |
| Windows Event Log on another computer | Application, source SyntinelRedAgent |
Windows Event Log entries.
- Source
YuviRed, information: "The YuviRed service has started", and "The YuviRed service has stopped" (event ID 4100) when it stopped because it was asked to. - Source
YuviRed, error: "YuviRed stopped at phase '<phase>': <error or authorization>". The phase is one from 1.1. This is the event to quote to support. - Source
YuviRedAgent(server) orSyntinelRedAgent(other computers): start and stop of the agent service, and the agent's own warnings. - System log, source Service Control Manager: a service failed to start, or ended unexpectedly.
The default agent log level records counts and errors, not the AI domains, program names or URLs it saw.