Skip to content

Releases

ReD-Pot 1.0.1

Choose a tag to compare

@github-actions github-actions released this 17 Nov 06:01

Bug-fix release from the 1.0.0 live test. No configuration or database change (rpot.json schema version 2).

Upgrading from 1.0.0: in place, keeping persona, baits, admin token, enrolled sensors and the database: on the
manager and on every remote sensor, git pull (or check out v1.0.1), then sudo ./install.sh update (--yes for
unattended runs). See "Upgrading" in docs/INSTALL.md. Older versions (v0.3.x / v0.4.x): uninstall, then install.

Fixed

  • Network map / console panels could fail with HTTP 500 and incoming events could be rejected under concurrent
    load.
    The console used one database connection for every thread, and the read paths (network map, world map,
    summary, incidents ...) ran on it while events were being stored. Overlapping use broke statements
    (InterfaceError: bad parameter or other API misuse), and an event batch answered with HTTP 500 could already be
    partly stored, so the decoy's retry stored some events twice. A batch is now stored completely or not at all. Reads now use their own
    read-only connection per thread (WAL: they see the last committed data and never block event storage); writes
    keep the single connection under the lock. The network map and world map APIs also share one computation
    between identical requests and reuse it for up to 5 s, so many open screens no longer pile up 24 h scans (their
    time range is rounded up to the UI's presets: 15 min, 1 h, 24 h, 7 days, 30 days, 90 days).
  • Network map: HTTP 500 "database or disk is full" on busy managers. The 24 h network map sorted every event of
    the range through an SQLite temp file (about 19 MB at 350k events a day), and the console's /tmp is a 16 MB tmpfs.
    SQLite's temp files now go to the data volume (SQLITE_TMPDIR=/var/lib/rpot/tmp, the same as the 1.0.0 hotfix),
    and the map query is bounded: the newest 20,000 source/decoy pairs, with a note when more exist; internal-only views
    are filtered in the query itself.
  • Attack map, "Live external activity" missed L4 events. Internal sources were filtered out only after a 400-row
    window, so a burst of internal traffic (a vulnerability scan, our own attack test) could empty the external feed even
    right after an external L4. Internal and allow-listed sources are now left out before the limit. The newest 10 L4+
    events of the range always stay in the 40-row feed, so internet scanner noise (L2/L3) no longer pushes an L4/L5
    out. The feed follows the map's level selector (never below L2), and its caption shows the level. Its cost depends
    on the feed size, not on the number of sources: events are read newest first in pages and the scan stops when the
    feed is full or reaches the start of the time range (hard cap 25 000 rows per scan); source classification uses a
    bounded LRU cache.
  • IPv4-mapped IPv6 addresses (::ffff:192.168.1.10, from a dual-stack socket) were classified as external and
    never matched IPv4 entries of allowlist, internal_nets, wall_allow or the manager's decoy-network UI deny
    list. They are now treated as the plain IPv4 address everywhere a source is classified, and stored as plain IPv4
    (decoys send it that way, and the manager normalises incoming events and bait-document beacons too).
  • Installer: the admin token was printed twice at the end of a manager install (the rpotctl init line and the
    summary). It is printed once, in the summary, and kept in INSTALL-INFO.txt. rpotctl init run by hand still
    shows it.
  • Remote sensor on DHCP locked out after an address change. The decoy follows a new lease, the manager's
    sensor-API allow-list did not, and there was no supported way to change it. The sensor installer now warns when
    the manager link is a DHCP lease (static address or DHCP reservation needed; confirm, default no), the L4 "stopped
    reporting" alert of a remote sensor points at the fix, and the new install.sh set-access changes the lists (see
    Added).
  • A clock jump on the manager raised false tamper alerts. A forward step of the system clock (NTP after a snapshot
    restore, date -s) opened an L4 "stopped reporting — possible tampering" incident for every decoy. Decoy silence is
    now measured on the monotonic clock; a detected jump starts a grace period for decoys without a baseline and is
    logged once on the console output. Admin sessions still end on a large forward jump (cookie expiry is wall-clock
    time).

Fixed (security)

  • Firewall apply had an open gap and could fail open. firewall.sh apply removed every ReD-Pot chain first and
    hooked the new ones only at the end, while the console and the decoys kept running; Ctrl-C or a dropped SSH session
    during add-decoy / remove-decoy could leave them without rules. The rules are now swapped make-before-break: the
    new generation of the chains is built next to the one in effect and hooked only when complete, then the old one is
    removed; a failed build is dropped and the previous rules stay. After an interrupted run (even kill -9) the next
    apply recognises the complete generation by checking every hook point, never by name. apply ignores
    SIGHUP/SIGINT/SIGTERM; the installer ignores SIGINT/SIGHUP/SIGTERM from writing firewall.sh to applying it and
    runs it in its own session; when even the previous rules cannot be restored the stack is stopped before the rules
    are taken out, and a failed apply says so honestly when the host is not protected. ip6tables failures count only
    where IPv6 is on. The same script runs at boot.
  • Allow-lists widened silently. A network with host bits set (1.2.3.4/0, 192.168.1.10/2) was turned into its
    network (0.0.0.0/0 = anyone, 192.0.0.0/2). Such input is now refused with the network and the single-address
    form to type instead; interactive runs show the normalised list.
  • trust_xff: the bait-document beacon trusted X-Forwarded-For from any client and took its leftmost entry, so
    a client could name an allow-listed address and mute the alert. It is now honoured only when the TCP peer is in
    the new console.trusted_proxies list (empty = never, even with trust_xff), only its rightmost entry (the one
    the proxy adds) is used, the proxy is recorded (data.via_proxy), and a source taken from it is never matched
    against the allow-list. Off by default; the installer never sets it.
  • Allow-lists and ports read back from installer.json are validated again before they reach firewall.sh;
    installer.json and INSTALL-INFO.txt are written atomically and never through a symlink.

Added

  • sudo ./install.sh update: in-place update of a manager or remote sensor (1.0.0 → 1.0.1, also from a 1.0.0
    manager running the hotfix image rpot:1.0.0-hotfix1, which it supersedes). It refuses anything
    but a supported version step of the Community Edition (and says "already up to date" on 1.0.1), makes a backup
    first (database with the sqlite backup API, configuration, compose file, firewall.sh, unit, with owners and
    modes; /var/backups/redpot/<time>-<kind>-<version>/, root only), gets the new image the normal way (verified
    download or local build), rewrites compose, unit and firewall.sh (manager), restarts and checks that the console
    answers and the decoys run. On any failure (any error, also a dropped SSH session) it restores the previous image,
    configuration (owners and modes kept), unit and firewall rules, reports "rolled back" only once the service runs
    again, and otherwise says exactly what is wrong. An update cut off by a power loss is offered for rollback on the
    next run (--resume-rollback for unattended runs; refused if the files changed after it); until then add-decoy,
    remove-decoy and set-access refuse. Uninstall removes the update backups (kept with --keep-data).
    Signals are ignored during the critical part. The image now carries an io.redside.edition label.
  • sudo ./install.sh set-access: change the web UI, sensor API and host SSH allow-lists of a manager install
    without reinstalling (persona, baits and admin token stay). Interactive with the current lists as defaults, or
    --yes --allow-sensor 192.168.1.0/24 (only the named lists change). It applies the install's exposure checks: on
    an address a decoy shares, a list open to anyone or containing the decoy's own LAN is warned about and needs
    confirmation (default no); with --yes it is refused unless --accept-exposure is given. It runs the firewall
    preflight, rewrites firewall.sh and swaps the rules in (see above); if that fails, the previous firewall.sh
    stays on disk and installer.json is unchanged. Your current SSH client always stays allowed; open web UI /
    sensor API connections from removed addresses are cut when the conntrack tool is installed (otherwise they stay
    until they close); host SSH sessions are never cut automatically; unchanged lists change nothing.
  • set-access warns when the new sensor-API list leaves out an enrolled remote sensor's address (it would be
    refused and go offline); interactive runs ask to confirm (default no).
  • The remote-sensor relay logs one line (at most once a minute per kind) instead of full tracebacks while the
    manager is unreachable.
  • add-decoy onto the management interface runs the same exposure checks (and needs --accept-exposure with
    --yes for an exposed list).

Changed

  • Installer, web UI allow-list: when the web UI allow-list is restricted (interactive or --yes), the installer
    warns that bait documents only report from addresses allowed to reach the console: the bait Word documents call
    home to the web UI port. This is how the Community Edition works; README ("Community Edition limits") and
    docs/INSTALL.md say the same. A separate beacon endpoint is planned for ReD-Pot Enterprise.
  • "Live external activity" caption shows the level filter and that the newest L4+ events are always kept.
  • Docs: remote sensors need a static address or a DHCP r...
Read more

ReD-Pot 1.0.0

Choose a tag to compare

@github-actions github-actions released this 13 Nov 19:09

First release of ReD-Pot Community Edition (MIT), built from v0.4.0. It keeps every decoy role, the baits and
honeytokens, scoring, the dashboard and remote sensors, and sets the Community Edition limits: at most 3 decoys, one
decoy per network interface on that interface's own IP, a fixed persona and bait set, no outbound alerting and no
offline replay. See "Community Edition limits" in the README.

Upgrading: fresh installs only. Uninstall v0.3.x / v0.4.x (sudo ./install.sh uninstall), then install 1.0.0
(new persona, baits and admin token; remote sensors join again). Commands that change an existing install refuse a
configuration made by an older version.

Added

  • Export: Incidents → Export JSON downloads incidents and alert records for a date range (whole UTC days, To
    included, at most 366 days) as one JSON file with a "complete": true marker. Admin sessions only; one export at a
    time; larger than 200 MB is refused with the counts instead of being cut off.
    The same from the command line: sudo ./install.sh export --from D --to D [-o FILE] [--force] (file mode 0600, never follows symlinks, --force replaces only a regular file), and inside the
    console container rpotctl export (JSON on stdout) and rpotctl slots (decoy slot counts).
  • Decoy limit (3 decoys: local decoys + remote sensors + pending joins), enforced by the installer, rpotctl,
    enrollment, the database and the sensor API. The dashboard shows "n of 3 decoys", disables Add a sensor when full
    and explains why; pending joins can be cancelled, which frees the slot at once. Revoking a remote sensor frees its
    slot; a local decoy frees its slot only with install.sh remove-decoy. A full manager answers a valid join with
    "decoy limit reached" and the token stays unused.
  • Edition label in the dashboard header; edition in /api/summary, slot counts in /api/fleet.
  • sensor.gap note: after an outage a remote sensor reports how many events it could not deliver (count and first/last
    timestamp only), shown on that decoy's incident.
  • Alert records are stored in the database (alert_records) as the source for Export, whether or not alerts.json
    is enabled. They follow the event retention.
  • tests/lab_scenarios.sh setup|attack|all|down builds a 3-decoy bridge lab (FS01, DC01, WEB01) and runs the simulated
    workstations; DRY_RUN=1 prints the commands.
  • Release pipeline (GitHub Actions): CI on every push and pull request (tests on Python 3.12/3.13, CE gate, shell
    syntax, dependency audit). A version tag builds the image for linux/amd64 and linux/arm64, publishes
    ghcr.io/redsidesecurity/red-pot:<version> with provenance and SBOM, checks the pushed image's source-hash label
    against the installer's src_hash(), then creates the GitHub Release from this changelog.

Changed

  • SSH decoy: an open interactive shell is no longer closed when no further channel is opened within 60 s.
  • Connection floods (pentest F-02): one decoy-wide connection budget for every TCP service, SMB and HTTP included
    (250 concurrent, 20 per source IP, thread ceiling 320 under the container's 384); SMB sessions without a named logon
    idle out after 60 s. A listener no longer dies when the decoy runs out of threads: that client is dropped and the
    listener keeps serving. Sources that hit a cap raise a level-3 conn.flood event (once a minute per source).
  • SMB named-pipe flood (pentest N-3): SMB named pipes (\srvsvc, \wkssvc) no longer start a thread each; RPC
    requests are answered in the SMB session's own thread. Every open pipe is charged to the decoy's connection budget:
    at most 8 per SMB session, 16 per source IP, 128 per decoy, and none once the decoy is within 32 threads of its
    thread ceiling (a pipe is refused before a new client is). A refused open gets STATUS_INSUFFICIENT_RESOURCES; the
    session keeps working. Idle pipes close after 60 s (null session) or 300 s (named logon), and all of a session's
    pipes close with the session. The SSH decoy allows 10 channels per connection (OpenSSH MaxSessions), charged the
    same way.
  • Resource-exhaustion detection: a source that hits its own pipe cap raises smb.pipe_flood (level 4); its own
    SSH channel cap, or holding all its connection slots while the decoy is full, raises resource.exhaustion (level 4).
    When the decoy-wide thread or connection cap refuses a client, the manager opens or raises a level-4 decoy incident
    "Decoy NAME saturated - possible DoS". That incident names the top holders (source IPs and their counts), not the
    refused client, and always writes an alert record when it opens or rises (otherwise at most one per decoy per 5 min).
    Flood events now carry the total refused since the flood started (refused), the count since the last report
    (new) and duration_s. The first refusal is reported at once and a summary follows within 60 s.
  • Decoy service health: every 30 s each decoy checks its listeners; a dead or unbound one is restarted and reported
    (sensor.service_down / sensor.service_up). The manager opens a level-4 "service down - possible DoS" incident on
    that decoy and /api/sensors shows services_down and per-service health; the dashboard marks a down service in red.
  • Persona: the installer asks only the AD/DNS domain and the industry. Company name, NetBIOS name and mail domain
    are derived from the domain and shown on the review screen. Users come from a larger built-in name pool (70 first
    names × 72 last names), 14 picked per install by a random seed, which also picks the username format. Internationalised
    (xn--) domain labels are refused with a hint. --domain is required with --yes (no default domain).
    The NetBIOS name is the first domain label without hyphens (corp.ararat-bank.am → ARARATBANK); when that is
    longer than 15 characters, its first word (northwind-traders-intl.com → NORTHWIND). Before: ARARAT-BANK,
    or a 15-character cut such as REDSIDE-SECURIT (pentest F-04).
  • Fewer honeypot tells (pentest F-04, F-05):
    • Bait files are dated per file at office-hours times with their own minutes and seconds (before, every file carried
      the install time's seconds). .docx baits have Word-like metadata: author and editor from the staff list, created
      and modified dates matching the file date, app.xml, and ZIP entries dated like Word's (no install time).
    • The DC's LDAP directory looks like a real Server 2019 domain: about 125 objects, including the default containers,
      CN=Builtin with the builtin groups, the default domain groups, the Default Domain (Controllers) Policy GPOs,
      a company OU tree with department OUs and groups, and one workstation per employee (also known to the KDC).
      Every memberOf points to an existing object, member matches it, and the primary group is not listed.
    • A null session can no longer list shares (Server 2019 default). Logged-on bait users still can.
    • The NetBIOS node-status reply carries the decoy interface's real MAC address instead of a made-up one.
  • One decoy per interface: each decoy answers on its interface's own IP; the address question is gone. The
    installer offers each interface once and allows at most min(3, usable interfaces) decoys. A 1-NIC server holds one
    local decoy, which shares the management IP (host SSH moves to a random port). --no-ssh-move with a decoy on the
    management interface is an error. Interfaces with unusual names are listed as Not usable.
  • add-decoy offers only interfaces without a decoy, checks the decoy count with the running manager first, and needs
    the manager to be running.
  • Remote sensor install: no placement or address question; the decoy answers on the IP of the chosen decoy interface.
    On the manager-link interface host SSH moves (and the decoy follows DHCP there); on another interface the decoy binds
    that IP and host SSH is pinned to the manager-link IP. The decoy-interface firewall policy applies on sensors too.
  • Manager unreachable: a sensor retries the batch in flight in memory for up to 60 s; after that, events go only to
    a write-only local log undelivered-<NAME>.jsonl on the sensor (mode 0600, RPOT_UNDELIVERED_MAX_MB, default 50) and
    are never sent later.
  • Sensor timestamps older than 15 minutes (was 7 days) or more than 5 minutes in the future are replaced by the
    manager's time; the original is kept as data.sensor_ts, and the manager log warns once per sensor about clock skew.
  • rpot.json format version 2. alerts keeps only min_level (the alerts.json threshold).
  • rpotctl init takes --domain (required) and --profile; --net-mode is host (production,
    --sensor NAME=role@IP%IFACE) or bridge (isolated lab). At most 3 --sensor entries.
  • The version (1.0.0) has a single source, rpot/__init__.py; the installer, rpotctl and the image tag read it.
  • DC DNS fingerprint: the domain controller decoy's DNS now answers like a default Windows Server DNS: CHAOS
    queries (version.bind) get NOTIMP with no version, other answers have RA set (header bit only; the decoy still
    never recurses), and non-QUERY opcodes (DNS STATUS) get a 12-byte NOTIMP. nmap -sV reported it as
    "NLnet Labs NSD"; it now reports "Simple DNS Plus" (OS: Windows), as it does for most real domain controllers.
  • DNS zone apex: the DC and Linux DNS decoys answered NXDOMAIN for the domain itself (SOA, NS, A) and for
    _msdcs / _tcp while names below them resolved, and negative answers had no SOA. The DC now serves the domain
    zone and an _msdcs zone with Windows defaults (SOA primary = the DC, hostmaster.<domain>, 900/600/86400/3600),
    NS and apex A = the DC, the Net Logon records (site, pdc/gc, DC and domain GUID records, DomainDnsZones /
    ForestDnsZones). The Linux DNS role gets a BIND-style SOA and NS. Empty names such as _tcp answer NOERROR with
    no data, NXDOMAIN/NOD...
Read more