The latest Call Telemetry releases and improvements.

This maintenance release is available on the stable channel for appliances on the 0.8.6 line.
Upgrade behavior change: CallManager phone web scraping now honors the Enable phone query setting. That setting defaults to off, so after upgrading, CallManagers without phone query enabled no longer scrape phone web pages for neighbor (CDP/LLDP) and serial details. AXL, RIS and Extension Mobility discovery are unchanged. To keep phone-page enrichment, enable phone query on each CallManager.
No enhancements in this release.
Fix: End-user sync from CUCM no longer logs duplicate-insert errors when two sync batches update the same user at the same time. The synced user data was already correct; the conflicting write is now skipped instead of reported as an error.
Fix: Network page discovery no longer logs warnings when IPv6 is disabled or absent (such as 0/ prefixes), reserving warnings for genuinely malformed subnet values.
Fix: DIME CDR backfill classifies CUCM job not found responses as a distinct, terminal job-not-found outcome rather than retrying as a generic failure.
Fix: Watch list violation lifecycle seeding now links blocked call events to the violation, ensuring blocked call counts accurately reflect blocked attempts.
Fix: Missed-call deduplication is now scoped per organization, preventing missed calls in one organization from suppressing notifications in another.
Fix: CUBE details and updates are now restricted to the owning organization; previously a request could read or modify a CUBE belonging to another organization.
Fix: CUBE update requests now accept flat request bodies without a nested wrapper, returning a valid update response instead of an error.
Fix: Emergency location report routes now restrict operations to available actions, preventing errors when requesting individual report details.
Fix: System performance monitoring uses growth-based alerting for BEAM literal allocator usage, eliminating false-positive error-level alerts under normal memory allocation.
Security: Updated com.google.guava to 33.7.2 in the JTAPI sidecar, addressing CVE-2026-102554.
Trivy 0.74.0 scan of the shipped linux/amd64 images for 0.8.6.33. Scan date: 2026-11-09 (UTC). Trivy vulnerability database: 2026-11-08 (UTC). Counts are Critical and High findings in the image scan. The full per-container report is at https://docs.calltelemetry.com/releases/scans/0.8.6.33-trivy-scan.html. | Image | Tag | Digest | Critical | High | | --- | --- | --- | ---: | ---: | | `alertmanager` | `v0.34.2` | `sha256:39d8bb3966c694a124428a3a88fbcffd40bd069129ecd47c059f8a3f17c55f3b` | 0 | 2 | | `alloy` | `v1.21.3` | `sha256:8ab871206bd2d8f89277f53abffee5f452a293980788d52f85fa183a87a505a4` | 0 | 3 | | `appliance-mcp` | `1.2.3` | `sha256:ae31842cdf522e349e25f5a91f9c0b65519e8b0d658a5079b0f42e490660dbfe` | 0 | 0 | | `caddy` | `2.11.8-alpine` | `sha256:f6228111e5fafa1a5238258d0858e103f10599b8d1659a146e7726681a002bb9` | 0 | 3 | | `ct-media-go` | `1.5.14` | `sha256:e8401cffa3969fd6b50d8206dced5b1a2d1b67cdc4e7f4dda18c6e85f642a8d6` | 0 | 46 | | `ct-syslog-ingest-go` | `1.1.8` | `sha256:08f8a932a76a8c709ad42bb08b18d88a8e0c5572d07f87d53495f00ecc1db21e` | 0 | 0 | | `grafana` | `12.4.14` | `sha256:ae509d9ae8c1021ca32ec370a9ca975d40165eda9fccd1e1102d317ff8787cd7` | 0 | 4 | | `jtapi-sidecar` | `0.8.7-beta-e5680d4685a2` | `sha256:61134a54673f3ef0148d04de693829c25ea9d381f161509787250362eea884ad` | 0 | 0 | | `loki` | `3.7.9` | `sha256:0db216b630ae1f348a1584c67f3cead8b4c1fc839169a39501db664744158e28` | 0 | 1 | | `nats` | `2.15.2-alpine` | `sha256:823e7f1d6d8ffe3b30dbd73d4a06bbb939acf94929f7fdf224404da7a05bafb5` | 0 | 0 | | `node-exporter` | `v1.12.1` | `sha256:1b4e4438faca4dd7e001dd445d161a4a2091b0fededa84093b3a8dfeae1f1be0` | 0 | 16 | | `opentelemetry-collector-contrib` | `0.165.0` | `sha256:e40d61f10c9a617cca1ffd1be4d7e2b3a4b0f0001798ff07e58cabe82e618cea` | 0 | 0 | | `postgres-14` | `14` | `sha256:0cc1fe2fb53fba71f11dde77206fda6e04d3bf7bfb6869013d9219a7bbe4a4ff` | 8 | 63 | | `postgres-18` | `18` | `sha256:602a442f31124465cb0c51b0aee82c413fa6a4631e7080bb645c008a0bebd88f` | 8 | 63 | | `postgres-exporter` | `v0.20.1` | `sha256:ac5ec343104fae0e2d84a27bb8d69b38430a11910c5382cad85d478d2bab713e` | 0 | 19 | | `prometheus` | `v3.15.1` | `sha256:fa3302279710be2f21a8301ec17a8b4f2301fd829bd113839b3b6b02b6f191a2` | 0 | 2 | | `prometheus-nats-exporter` | `0.20.3` | `sha256:4b8ddc1c86fec360b5cd3d5e1fbc6fb52bdafd5805bba3901cc52926538af255` | 0 | 3 | | `seaweedfs` | `4.52` | `sha256:57877e32bb90073b37dc109b438ba66e84fd5ed9ced5f5c04f56444ef243ce59` | 0 | 1 | | `tempo` | `3.1.1` | `sha256:64e2fbd56d9c2b3412a4b430a58ea59e91a670b1e75f920a0c702213ad145044` | 0 | 2 | | `traceroute-go` | `1.4.11` | `sha256:922a1432ca71fa4c1b53faa792d8811ed01ae869fcbaf168b5ff68647dac5d1a` | 0 | 48 | | `vue` | `0.8.6-stable-035c35895cef` | `sha256:ba1e5399bfb8e7eb374a0593b7b044ac4ab6a05d9de3de213b8d4cb630c001c6` | 0 | 0 | | `web` | `0.8.6-stable-36bf85592569` | `sha256:7a4ef3adcb8625af8f7ce346b34000f33a2ca80a401eafab61bbcae976a1f95c` | 0 | 0 | The `web` image Trivy scan is listed above. A supplemental Elixir audit of that component also reported 3 Critical and 6 High findings, which are not included in the Trivy counts. Those findings are in the full report.
Update with:
sudo curl -fsSL https://get.calltelemetry.com | sudo sh -s -- update stable
This maintenance release is available on the stable channel for appliances on the 0.8.6 line.
Enhancement: Hide cached retries. Policy History has a Hide cached retries filter and a badge that marks repeat decisions served from the short-lived cache. Repeat decisions are no longer counted as new blocked calls in dashboards and reports.
Enhancement: Policy app changes in the activity log. Changes to policy apps now appear in the Policies activity log.
Enhancement: Caller and called device fields in CDR reports can be filtered and sorted, and the advanced filter lists only fields that support filtering.
Enhancement: One CIDR for emergency subnets. Create a subnet by entering a single CIDR value. The network type (IPv4 or IPv6) is detected automatically, and CSV imports accept a single CIDR column.
Enhancement: The Emergency Subnets page no longer shows a Discover from Meraki button or its license notice. Run Meraki network discovery from the Meraki Integration page, and set Meraki lookups and the Meraki token in E911 Settings.
Enhancement: CUBE call lists filter by outcome (All, Completed, Failed, Rejected, or Unknown) on the global and per-CUBE pages.
Enhancement: Teams Reporting data. Teams Reporting can sync call records, call and video quality, adoption activity, PSTN usage, and license assignments, and its tables follow the same retention and cleanup settings as other report data. Call records, PSTN usage, and license sync are opt-in per organization and stay off until enabled in Teams settings.
Enhancement: Teams events and alarms. Teams events are received through Azure Event Hubs, which is the only subscription path in this release. Subscription, Event Hub, and token-health problems raise alarms.
Enhancement: Teams emergency calls. Teams emergency calls are mapped to civic addresses and routed to the matching location for emergency workflows. Emergency workflows require the Teams connector license (Premium, Demo, or Beta).
Enhancement: Teams setup and credentials. Microsoft sign-in uses OAuth with a one-time exchange, and Graph access separates delegated and application permissions. Azure OAuth and Event Hub settings are entered on the Teams Connector page. Teams Reporting runs only for organizations that have Microsoft credentials and have enabled call records, and Teams credentials and tokens are no longer written to logs.
New feature: Job Scheduler console. A console lists background jobs for your organization, with a separate view of appliance background jobs and each job's most recent observed run.
Fix: Blocks from app-based decisions (blacklist, watch list, duplicate, and block app) stay blocks. Later default rules no longer turn them into permits, and external risk classifications are honored when present.
Fix: Policy History shows a duplicate-blocked call as a duplicate when no app signal is recorded, instead of a later default permit.
Fix: Opening a new rule no longer sends requests for a rule that does not exist yet or shows a premature error banner.
Fix: Rule render and preview requests are now restricted to the requesting organization's own policies and rules.
Fix: Report filters for contains, does not contain, begins with, and ends with work as expected. Some text filters were previously rejected with an error or matched only exact values.
Fix: Calls that never connected no longer show a 1969 or 1970 connect date, and they stay in reports that filter by connect time.
Fix: Saved reports with an invalid definition return a validation error (HTTP 422) instead of a server error (HTTP 500).
Fix: Custom CDR History reports with a field condition run instead of returning a server error, and CDR History date filters accept calendar dates.
Fix: Scheduled PDF reports limit their data to what the PDF format can hold, so they deliver instead of failing with a row-limit error.
Fix: Phone, end-user, subnet, and switch report filters apply the values you enter, and phone joins match device names regardless of letter case or surrounding spaces.
Fix: Phone registration health and trend reports are no longer denied to Essentials-licensed organizations when run through AI report actions.
Fix: Dashboard call volume and the reputation timeline follow the selected time range and count the same calls.
Fix: Cards and charts that showed randomly generated figures (greeting usage, workflow status, and CUBE mini metrics) now show appliance data or an empty state.
Fix: Retrieving a single emergency (E911) address returns the address, or a 404 when it does not exist, instead of a server error.
Fix: User sync from CUCM no longer aborts when a CUCM response begins with a byte-order mark or arrives encoded as UTF-16.
Fix: Phone prefix discovery works again for phone IPv6 addresses written in compressed form.
Fix: CUBE event logs load on first open. The Registration Analysis dialog shows a real success rate and opens from the Current Status chip.
Fix: Clicking a CUBE name opens its metrics page. The CUBE stats strip no longer shows zero events and failures above a populated table, and search highlights in CUBE XML views display matches.
Fix: CUBE authorization returns a valid call route for permit decisions that use a called-number modifier.
Fix: A network error during token renewal no longer signs you out. Only a renewal the server rejects does.
Fix: After an upgrade, the web interface confirms your session with the server before it shows protected pages, instead of showing a stale logged-in view.
Fix: Bulk contact imports that update existing contacts match the row from the same source, so duplicate phone numbers across sources no longer fail or update the wrong contact.
Fix: Appliance updates and ct appliance export work on appliances running 0.7.x, and ct appliance import from a 0.7.x archive leaves the restored appliance able to connect to its database.
Fix: Fresh 0.8.6 installs no longer print a database role warning on every update, and ct show-tech bundles collected from the web interface no longer list Docker sections as errors.
Fix: Repeated update-check failures back off instead of retrying on every request, so an unreachable release channel no longer stalls the update screen.
Fix: Emergency disk-pressure cleanup on containerized appliances measures free space for the whole host, so it runs when the host is actually low on space.
Fix: Data retention measures each new purge target on its first pass, and a measurement error no longer stops purging for an organization.
Fix: Data retention purges confirm their target table before deleting rows, so a mismatched retention entry cannot remove records from another table.
Fix: Background NATS bucket reconciliation completes on appliances with JTAPI disabled instead of failing until it gives up.
Fix: Organizations on the Community tier keep the Community feature set (beta, tags, and numbers) when their subscription record is inactive, and database size refreshes are recorded in the audit trail.
Update with:
sudo curl -fsSL https://get.calltelemetry.com | sudo sh -s -- update stable
This maintenance release is available on the stable channel for appliances on the 0.8.6 line.
Enhancement: Policy History follows your time zone. Days in Policy History now start and end at midnight in your organization's time zone instead of UTC. The default 24-hour view also moves forward each time you change a filter, rather than staying fixed at the moment the page opened.
New feature: Tab completion for the ct command. Press Tab to complete commands, options and app names, with a short description for each option. Updating installs it for bash and zsh and reinstalls it if it goes missing, and ct doctor reports whether it is working.
Fix: Appliances with very large policies (1,000+ rules) stay up under load instead of running out of memory.
Fix: Call routing keeps working while live debugging is on, and post-call actions are no longer skipped.
Fix: Policy History charts place calls in the correct hour and day, and filtering by app action finds the matching calls.
Fix: Calls without caller ID display correctly in policy analytics, and the TDoS overview updates when you change filters.
Fix: Large uploaded CDR and CMR files import reliably, and faster.
Fix: JTAPI phone control keeps running when an optional JTAPI feature fails to start.
Fix: E911 bulk address imports show which rows failed, and a switch can only be assigned a saved address.
Fix: CUCM log collection resumes after an interruption, and AXL query errors explain what went wrong.
Fix: ct health --wait waits until the appliance is actually ready, and ct app up is safe to run on a stack that is already running.
Fix: LDAP sign-in says clearly when no matching user exists. The update check stops flooding logs with certificate errors, and the update screen says when it is showing fallback release information.
Security: LDAP sign-in safely handles special characters in user names.
Security: SMTP passwords are no longer returned by the API or shown in settings.
Security: A development certificate key is no longer shipped in the web image.
Security: Logs and error messages no longer include sensitive values.
Security: Resolved CVE-2026-14456 (High): updated libcrypto3 3.5.7-r0 to 3.5.8-r0 and libssl3 3.5.7-r0 to 3.5.8-r0 in the prometheus-nats-exporter image.
Update with:
sudo curl -fsSL https://get.calltelemetry.com | sudo sh -s -- update stable
This maintenance release is available on the stable channel for appliances on the 0.8.6 line.
Enhancement: Multi-number search. The Calling and Called filters accept several numbers and match any of them. This works in CDR Reports, CDR History, Alerts, Missed Calls, Summary by Extension and Policy History, and international numbers keep their leading +.
Enhancement: From reputation to the calls behind it. Risk levels and blocked categories on the reputation and blocking cards now open Policy History. The matching filters and date range are applied as chips you can remove.
Enhancement: Redirected calls in block reports. Policy analytics count a redirected call as blocked and show how many of those blocks were redirected. Block totals stay comparable with earlier releases while showing how many callers reached a destination instead of being rejected.
Enhancement: AI providers and reasoning control. AI settings support nine providers natively: OpenAI, Anthropic, Google, DeepSeek, xAI, neuralwatt, OpenRouter, Azure OpenAI and Ollama. A reasoning-effort setting, from none to max, controls how much reasoning each request uses.
Enhancement: Phone dashboard. Phone scanning is redesigned, with scan controls, progress and scan history together on the phone dashboard. The remote-control phone view keeps the handset at full size with a collapsible details panel, and fits small screens without panning.
Enhancement: E911 placement at a glance. Subnet and switch pages show how many phones each one holds. Select the count to see a diagram of those phones and their switch ports, and open any phone's details from it.
Enhancement: Job Scheduler in the sidebar. The Job Scheduler is now listed under Settings in the left sidebar.
New feature: Blocked-call treatments. Choose what a blocked caller experiences for each block reason: reject the call, optionally with an announcement, or redirect it to your organization's default number or a number you choose. Ten reasons are covered, including blacklist, block rules, watch list, TDoS protection, device call quota and duplicate calls, and call tests and Policy History show which treatment a call received. Redirects require an Advanced license.
New feature: Duplicate call suppression. A policy can block a caller who calls the same number more than a set number of times within a time window. Turn it on in the policy's Rate Limiting settings and set the threshold and window; whitelisted callers are always allowed.
New feature: Contact tags. Tag local contacts in bulk or during CSV import, and use the tags in policy rules, blacklists and whitelists. Each tag maps to Permit, Redirect or Block, and the contact test shows the resulting action and the matched tags.
New feature: Caller names from your own PostgreSQL database. Sync contacts from a PostgreSQL directory into Local Contacts on a schedule, so calls show the names your organization already keeps. A guided editor with a schema browser, test run and field mapping builds the query, and the Inbound Caller ID Test checks any number.
New feature: SMTP History. See every email the appliance sent, including test messages, scheduled reports and password resets, with its delivery result. Open it from the Email tab in Settings to filter by recipient, status, source or date, and to preview or download attachments until they expire.
New feature: Disk-pressure alerts. When the appliance has to free disk space in an emergency, your alert destinations (email, SMS, Teams, Webex or webhook) receive one notice per incident. The notice reports the free space left and the records removed, and it does not depend on SMTP being configured.
New feature: Send Logs to Support. Send an encrypted log bundle to support in one step from Settings → Troubleshooting, with an optional note and no support case needed. You get a receipt ID to quote, and downloading the logs as a ZIP is still available.
Fix: Redirect destinations are checked when you save them, throttled calls stay blocked, and CUBE reports the right reason when it releases a soft-blocked caller.
Fix: Policy History stays fast on long time ranges, shows when its data was last refreshed, filters by reputation, and marks exports that were cut off.
Fix: Report delivery shows, for each destination, whether the report was sent and in which format.
Fix: New appliances reserve emergency disk space automatically, and setup and diagnostics work when optional services are turned off.
Security: Resolved 31 vulnerabilities (1 Critical, 30 High) in the alertmanager image by updating 6 packages.
Security: Resolved 50 vulnerabilities (2 Critical, 48 High) in the alloy image by updating 14 packages and removing 4 packages.
Security: Resolved 16 vulnerabilities (16 High) in the caddy image by updating 6 packages.
Security: Resolved 10 vulnerabilities (3 Critical, 7 High) in the ct-media-go image by updating 3 packages.
Security: Resolved 30 vulnerabilities (2 Critical, 28 High) in the ct-syslog-ingest-go image by updating 7 packages.
Security: Resolved 66 vulnerabilities (7 Critical, 59 High) in the grafana image by updating 23 packages and removing 1 package.
Security: Resolved 8 vulnerabilities (3 Critical, 5 High) in the jtapi-sidecar image by updating 5 packages.
Security: Resolved 40 vulnerabilities (3 Critical, 37 High) in the loki image by updating 9 packages.
Security: Resolved 18 vulnerabilities (18 High) in the nats image by updating 5 packages.
Security: Resolved 32 vulnerabilities (1 Critical, 31 High) in the node-exporter image by updating 5 packages.
Security: Resolved 60 vulnerabilities (4 Critical, 56 High) in the opentelemetry-collector-contrib image by updating 14 packages and removing 9 packages.
Security: Resolved 26 vulnerabilities (1 Critical, 25 High) in the postgres-exporter image by updating 3 packages.
Security: Resolved 36 vulnerabilities (2 Critical, 34 High) in the prometheus-nats-exporter image by updating 9 packages.
Security: Resolved 42 vulnerabilities (4 Critical, 38 High) in the prometheus image by updating 8 packages and removing 1 package.
Security: Resolved 5 vulnerabilities (5 High) in the seaweedfs image by updating 4 packages.
Security: Resolved 43 vulnerabilities (2 Critical, 41 High) in the tempo image by updating 11 packages.
Security: Resolved 17 vulnerabilities (17 High) in the traceroute-go image by updating 3 packages.
Update with:
sudo curl -fsSL https://get.calltelemetry.com | sudo sh -s -- update stable
This maintenance release is available on the stable channel for appliances on the 0.8.6 line.
Enhancement: ct doctor health checks on the cli.
Enhancement: CUBE XCC call logging now consolidates records into one event row per SOAP transaction, reducing disk footprint by ~22% per call.
Fix: Fixed a database maintenance issue where scheduled purge routines targeted webhook configurations instead of execution logs.
Fix: CUBE call control is now decoupled from event logging so that a database logging failure cannot interrupt live call processing.
Fix: Corrected CUBE authorization metrics on gateways configured with reject codes, restoring proper health badge calculations from degraded to healthy.
Fix: Addressed schema and enum mismatches in CUBE event logging that caused event rows with integer disconnect codes to be silently dropped.
Fix: Scheduled report destinations in the web UI now retain enabled/disabled toggle states and properly match schedule UUIDs.
Fix: Resolved a startup race condition where the CDR event subscriber could encounter an error before the event bus was fully initialized.
Fix: Hardened diagnostic exports and ct support-upload operations with session resume support and improved progress reporting.
Security: Patched container runtime dependencies (Chromium and libexpat1), resolving 53 security vulnerabilities (CVEs) across AMD64 and ARM64 appliances.

This maintenance release is available on the stable channel for appliances on the 0.8.6 line. It hardens appliance update diagnostics, watch list event processing resilience, policy analytics performance, and alarm audit durability.
No enhancements in this release.
Fix: Local update diagnostics report. Adds a local ct diag update report to capture appliance service restart failures during upgrade. The report is saved locally on the appliance for troubleshooting and is not uploaded automatically.
Fix: Watch list event accumulator resilience. Resolves ETS table key mismatches and orphan cleanup handling in WatchListBlockEventAccumulator. Database errors during event processing are handled cleanly without crashing the accumulator process.
Fix: Policy analytics query performance. Optimizes policy analytics total count queries, significantly speeding up dashboard analytics loading on large datasets.
Fix: Alarm telemetry and audit durability. Hardens alarm dismissal advisory locking, audit log duration tracking, and notification dispatcher retry resilience.
Fix: Emergency disk pressure buffer flush. Adds an emergency flush mechanism for curri_events under high disk pressure, preventing memory exhaustion and database lockouts when appliance storage is constrained.
Fix: Scheduled report destinations backend. Adds backend API foundation for delivering scheduled reports to configured external destinations.
Fix: Storage DNS preflight. Adds S3 and SeaweedFS DNS preflight resolution and literal allocator headroom checks to prevent storage connectivity timeouts during startup.
Update with:
sudo curl -fsSL https://get.calltelemetry.com | sudo sh -s -- update stable
This maintenance release is available on the stable channel for appliances on the 0.8.6 line.
This stable release hardens registration reporting for clusters where CUCM nodes return partial, fixes CUBE/XCC dialogs, and fixes some database purging, background jobs, and public dashboard widgets.
Fix: RIS registration and collection fixes
Fix: CUBE/XCC application and dialog URLs fixed from 4001 to 80
Fix: Data purging fixes
Fix: Public dashboard Widget fixes
Fix: Pressing Enter in the Add Peer IP field adds the peer instead of submitting the dialog.
Fix: CMR materialized-view and job fixes.
Update with:
sudo curl -fsSL https://get.calltelemetry.com | sudo sh -s -- update stable
This maintenance release is available on the stable channel for appliances on the 0.8.6 line.
No enhancements in this release.
Fix: Scheduled data purge now completes on large databases. Purge previously scanned each table in full and could time out before deleting anything, so retention silently stopped enforcing and disk usage kept growing. This release adds the missing indexes and the purge job runs to completion.
Fix: CUBE XCC charts now render in the timezone selected in the UI instead of UTC, so concurrent calls, call counts, and trunk activity line up with the rest of the interface.
Fix: Top Caller history drill-down now matches the totals shown in the rollup, including calls of five seconds or less.
Fix: Peer-health MOS now shows a gap where no gradable audio was measured, instead of reporting it as a zero score.
Fix: The phone dashboard Refresh button now refreshes KPIs and statistics along with the phone list.
Fix: Upgrades now check free disk space before starting. The index work in this release temporarily needs room beyond the database itself, and an appliance without enough space is stopped before any change is made rather than failing partway through.
Fix: A migration that cannot complete now waits before retrying instead of retrying continuously, so a single problem cannot consume disk through repeated attempts.
Update with:
sudo curl -fsSL https://get.calltelemetry.com | sudo sh -s -- update stable