News:

We really need your input in this questionnaire

Main Menu

NetXMS 6.2 patch release 2

Started by Victor Kirhenshtein, July 28, 2026, 11:04:02 AM

Previous topic - Next topic

Victor Kirhenshtein

We just published patch release 2 for version 6.2. Important changes since release 6.2.1:

-   Added subagent for reading electricity grid data (including carbon intensity) from ENTSO-E Transparency Platform
-   Added subagent for reading weather and solar forecast data from open-meteo.com
-   Device configuration backup support for HPE ProCurve, HPE Comware, and Aruba switches
-   New "scripted gauge" dashboard element
-   Agent checks available disk space before starting upgrade
-   Number of attempts to deploy a software package is now limited
-   Agent connect timeout on the server side is now configurable
-   Number of crash dumps kept on disk can be limited
-   File manager improvements: folders expand on double-click, view state preserved on manual refresh
-   Fixed remotely triggerable stack buffer overflow, heap over-read, and memory leak in NTCB module packet parser
-   Fixed authenticated memory corruption in client session (CMD\_MODIFY\_NODE\_DCI)
-   Fixed agent communication freeze caused by socket descriptor reuse race in communication channel close
-   Fixed thread pool starvation that could make agent unresponsive when all worker threads are blocked
-   Fixed deadlock in logwatch subagent
-   Fixed creation of duplicate ::/64 subnets on every configuration poll
-   Updated bundled libcurl in Windows installers to 8.21.0 (fixes CVE-2026-8927)

And full list of closed issues:

-   #3277 (Limit number of attempts to deploy a software package)
-   #3388 (NXSL DateTime: fields not normalized after change; out-of-range month breaks formatting)
-   #3390 (Web UI: "Restart agent" button in agent restart notification throws "display must not be null")
-   #3392 (logwatch subagent deadlock)
-   #3393 (Unable to query Table DCI columns when using source node override)
-   #3394 (Make interface class SetPeer method available through API)
-   #3396 (Expand folders in File Manager on double-click)
-   #3397 (ENTSO-E Transparency Platform subagent for grid carbon-intensity data)
-   #3398 (Update bundled libcurl.dll in Windows installers to 8.21.0 (CVE-2026-8927))
-   #3400 (Map tile cache not invalidated when tile server URL changes (stale/missing tiles))
-   #3401 (File manager view should preserve state on manual refresh)
-   #3402 (Package deployment manager loads all jobs into memory at startup regardless of status)
-   #3403 (nxmc: external browser never opens on macOS)
-   #3405 (nxmc: unhelpful error when client and server protocol versions differ)
-   #3407 (ai\_task\_execution\_log is never cleaned by housekeeper (unbounded growth))
-   #3408 (open-meteo.com subagent for weather/solar forecast data)
-   #3410 (Default zone (object ID 4) is never persisted, loadCommonProperties() fails on every startup)
-   #3411 (SQL error when creating new node on Oracle DB)
-   #3414 (Agent connections freeze until server restart due to socket descriptor reuse race in comm channel close)
-   #3419 (Summary table selector should show menu path if title is empty)
-   #3420 (Per-tag debug level specified from command line should override level specified in config file)
-   #3421 (Add device configuration backup support to HPE/Aruba drivers (ProCurve, Comware, ArubaOS))
-   #3424 (nxsl: Data stack underflow)
-   #3426 (Create limit for crash dumps)
-   #3427 ("Cannot start Java application" when exiting Management Client on Windows)
-   #3430 (Missing "Go to object" from object query table as a Dashboard element)
-   #3431 (Make agent connect timeout configurable)
-   #3433 (Check free space on file system before agent update)
-   #3436 (Thread pool does not grow when all workers are blocked; agent becomes unresponsive on COMM pool starvation)
-   #3438 (False "system restart detected" in status poll forces interface resync, bypassing disabled configuration poll)
-   #3440 (FindRemoteAccessPoint() called with chassis ID subtype instead of chassis ID)
-   #3441 (Node can be reported as its own root cause when interface peer resolves to the node itself)
-   #3443 (Config import silently turns EPP rule into catch-all when referenced event template is missing)
-   #3445 (SPARC: object index forEach stops after first element (enumeratorWrapper discards callback result))
-   #3446 (Client session: indeterminate mapId/mapIndex/mapCount in CMD\_MODIFY\_NODE\_DCI (authenticated memory corruption))
-   #3449 (Stop and restart of external subagent process does not stop loaded subagents correctly)
-   #3450 (Race condition in external subagent request ID generation)
-   #3451 (NTCB module: remotely triggerable stack buffer overflow, heap over-read and unbounded leak in packet parser)
-   #3452 (libethernetip: CIP byte order, unaligned reads, and truncated Identity serial number)
-   #3453 ("Hide in view mode" option missing for table DCIs)
-   #3455 (Scripted gauge dashboard element)
-   #3456 (FileDeliveryPolicy with empty content logs "XML parser error ... no element found" on every config poll)
-   #3457 (getPerfTabDCIList: getPerfTabSettings() != nullptr guard is always true)
-   #3460 (MikroTik driver advertises HostMib.Memory metrics that getMetric cannot serve)
-   #3461 (Network map layout changes are not saved until another action is performed)
-   #3462 (Incorrect layout option selection when opening a Network Map)
-   #3463 (Duplicate ::/64 subnets created on every configuration poll)
-   #3465 (nxmc: -server= cannot carry an IPv6 address with a port)
-   #3468 ("Use server time zone" option is not applied on management client startup)