SecurityApplication security

You patched your Cisco email gateway. One email can still take it over.

Attackers are getting root on Cisco Secure Email gateways with nothing more than an email. Cisco re-issued the 15.5 fix after exploit code appeared, and 16.0 boxes patched in September still have work to do.

Cohash Security Practice November 2026 13 min read 7 chapters 3-question check
part 37 of 37 · Platform releases, explained from scratch
01 Crafted email 02 Mail parser bad input trusted 03 Root ongateway 04 16.5 and ahunt upgrade and check

the ideaOne crafted email can reach root. Upgrade, then check nobody got there first.

after reading, you can
  • Tell which AsyncOS build each gateway needs, including the 15.5 fix Cisco re-issued
  • Check a gateway for signs of attack using Cisco's own commands, then confirm it off the box
  • Rebuild a gateway someone broke into, then replace every key and password on it

Someone can take over a Cisco email gateway by sending it one email, and nobody has to click anything. If you run these gateways yourself, get them to 16.5, then check whether anyone got in before you did.

in plain wordsCritical: exploited, act today

Someone can take over your mail filter by emailing it. Nobody inside has to click a thing.

Plenty of companies run a Cisco Secure Email Gateway in front of their mailboxes. It reads every message coming in and throws out the spam and malware. Two bugs in that reading step let a stranger run commands on it as root, the account that can do anything on the machine. Cisco says attackers are using both. Nobody on your staff has to open a thing, since the gateway does the reading. If Cisco hosts yours, it's already been upgraded, but you still need to find out whether Cisco has contacted you.

think of it like this

Think of a company mailroom where one clerk opens every parcel to check it's safe. These bugs are parcels built to take over the clerk. From then on the clerk works for someone else, and that clerk holds the keys to the rest of the building.

who should care
  • Cisco Secure Email Gateway owners
  • Teams running virtual gateways
  • Companies with clustered gateways
  • Cisco Secure Email Cloud customers
chapter 01

Why it matters to you

An email gateway is a bad machine to lose. It sits in the path of your mail, and it holds working credentials for the systems around it. Cisco's own upgrade notes list what's stored in its configuration: certificate private keys, LDAP bind passwords, DKIM signing keys, SAML certificate passphrases and outgoing SMTP passwords. Anyone with root can go after all of that.

9.8CVSS score Cisco gives each of the two exploited flaws
2Flaws Cisco says attackers are actively exploiting
0Workarounds Cisco offers for either flaw
400+Internet-exposed gateways Shadowserver tracked in mid-September; may include honeypots

Nobody has said publicly who's behind this or how many gateways were hit. Rapid7 reads the same-day KEVdefinitionCISA's Known Exploited Vulnerabilities catalog, the US government's list of bugs with evidence of real attacks. Federal agencies get deadlines from it. listing as a sign the SQL bug was used before Cisco disclosed it. CISA lists ransomware use as unknown for both flaws, so don't read that as a no. Rapid7 also said in its write-up that no public exploit code existed at the time. Cisco has since referred to exploit code available online, without saying where it is or what it does, so plan as if anyone can get it.

A firewall won't save you here. The gateway has to accept mail from the internet to do its job, so the attack arrives through the door it's meant to keep open. Cisco says the SQL flaw affects gateways regardless of how they're configured, physical or virtual.

how the attack worksHow the email attack works, step by step
  1. 01 A crafted email arrives

    The attacker emails any address your gateway handles. There's no login to get past, and nobody at your company ever opens it.

  2. 02 The gateway parses it

    AsyncOS reads the message as part of normal mail handling. The input checks in that parsing code are too weak.

  3. 03 Mail text becomes a command

    SQL hidden in the email gets executed as a database statement, when it should have been handled as plain data.

  4. 04 Commands run as root

    From there the attacker can run commands on the operating system underneath, with full control of the appliance.

  5. 05 Evidence can disappear

    Cisco warns that root access lets an attacker remove or hide the traces, so a clean log on the gateway proves very little.

Engineers will notice Cisco's sample search, COPY.*TO PROGRAM. That's the shape of PostgreSQL's COPY ... TO PROGRAM statement, which makes a database server run a shell command. Cisco doesn't name the database in its advisory, so treat this as a clue about what to hunt for. The second bug takes a different route. It's a path traversal in how the gateway handles attachments when File Reputation is on, and it lets an attacker write files and run commands as root. The Hacker News reports that the two sit in different parts of the software and that Cisco hasn't linked the attacks.

chapter 02

What Cisco fixed, and what's still open on 16.0

The advisories that matter

fixedCVE-2026-76461: SQL injection in email parsing

Exploited. Affects physical and virtual gateways regardless of configuration. A crafted email leads to commands as root. Fixed in AsyncOS 15.5.5-014, 16.0.4-302 and 16.5.0-780.

fixedCVE-2026-76519: file write through attachments

Exploited. You're exposed if File Reputation is enabled on a mail policy and Content Scanner Tools is older than 25.3.0.5525. Fixed in 15.5.5-014, 16.5.0-780 and that Content Scanner Tools version.

changed15.5.5-014 fix re-issued

Cisco updated the fix on October 31 after exploit code appeared online. Gateways on this build need Cisco Retro Remediation Engine 4.0.2 or later to count as having it.

fixedHardening release: five more CVEs

Cisco's own review turned up five grouped CVEs, four of them with a top score of 9.8. They hit gateways and Secure Email and Web Manager. There's no fixed 16.0 build, so anyone on 16.0 has to move.

tableWhat's left to do depends on your AsyncOS line. We put all three Cisco advisories in one table
You're onSQL injection (76461)File write (76519)Hardening releaseWhat to do
15.5 and earlier15.5.5-014, with the engine check15.5.5-014 (Cisco's row here reads 15.5)15.5.5-014Get to 15.5.5-014 at minimum, then plan 16.5
16.016.0.4-302No fixed build. Content Scanner Tools 25.3.0.5525 or laterNo fixed buildMove to 16.5. Update Content Scanner Tools today
16.516.5.0-78016.5.0-78016.5.0-780Run 16.5.0-780 or later
Secure Email CloudCisco upgraded to 16.5.0-780Fixed by CiscoCisco maintains itAsk whether Cisco flagged your tenant

Watch the build number when you compare notes. The Hacker News's September report prints the 15.5 fix as 15.5.5-0141, and that build isn't in Cisco's table. Cisco's advisory says 15.5.5-014. Check your gateways against Cisco's table, and don't copy a build number out of a news story.

How this unfolded

  1. Sep 14, 2026
    SQL injection disclosed

    Cisco posts the advisory and a hardening release together. CISA puts the flaw on KEV that day and gives federal agencies until September 17.

  2. Sep 17, 2026
    Cluster warning added

    Cisco warns that SSH keys used between cluster members can be stolen from a bad gateway. One compromised member now means restoring them all.

  3. Oct 14, 2026
    Second exploited flaw disclosed

    CVE-2026-76519 goes on KEV the same day, due October 17 for federal agencies. Fixing the first bug doesn't always fix this one.

  4. Oct 31, 2026
    15.5.5-014 fix updated

    Cisco reacts to exploit code that's now online, so anyone on the earlier 15.5.5-014 fix needs to look again.

  5. Nov 3, 2026
    Advisory revised

    Version 1.2 tells anyone on 15.5.5-014 to check the engine version. The build number alone won't tell you.

  6. Nov 5, 2026
    16.5.0-825 release notes

    Cisco's new notes list 15.5.5-014 as a starting point for 16.5.0-825. Those gateways now have a documented route up.

side by sideA gateway you patched in September against one that's actually finished
Patched in September
  • 16.0.4-302 closes the SQL injection only
  • You're still open to file write if Content Scanner Tools is old
  • Hardening release fixes are missing on 16.0
  • 15.5.5-014 may not have the October 31 update
  • Nobody has looked for signs of a break-in
Finished
  • Running 16.5.0-780 or later
  • Content Scanner Tools at 25.3.0.5525 or later
  • Someone searched mail_logs and system_logs on every cluster member
  • Someone actually read the firewall logs for odd traffic in or out
  • Result written down, with who checked and when
chapter 03

What to do now

What to do now, by who you are

Business ownertoday
  • Ask IT or your provider whether any Cisco email gateway handles your mail, and whether Cisco hosts it or you do.
  • Get the running version of each one in writing, along with the date someone checked for break-ins.
  • If Cisco hosts it, ask them straight out whether they've flagged malicious activity on your devices.
  • Someone finds signs of a break-in? Call the lawyer and the cyber insurer that day. Waiting usually costs more.
Email and IT adminstoday
  • List every gateway, including virtual ones and each cluster member. Note the build and the Content Scanner Tools version.
  • On 15.5.5-014, run retroscannerstatus. The engine has to show 4.0.2 or later.
  • On 16.0, run contentscannerupdate now, then schedule the move to 16.5.
  • Turn on automatic Content Scanner Tools updates if someone switched them off.
  • Save logs off the box before you upgrade, because the upgrade reboots the appliance.
Security teamthis week
  • Search mail_logs for COPY.*TO PROGRAM on every cluster member. Cisco calls that example non-exhaustive.
  • Search system_logs for System Security lines. Cisco says what to do about each message, so follow that.
  • Pull firewall logs for unexpected uploads from the gateway and downloads from known-bad addresses.
  • Turn on Snort rules 3:67109 and 3:67110 if you run Snort in front of the gateway.
Federal teams and contractorsnow, overdue
  • Both KEV due dates have passed, and both entries are marked for forensic triage.
  • Collect evidence before patching. CISA's guidance warns that patching may destroy artifacts.
  • Contractors: read your contract, since BOD 26-04 reaches you only where the contract says so.
gateway-checks (Cisco's commands, example only)example output
# Attempted SQL injection: run it on every cluster member. Any hit may be bad news
cisco-esa> grep -i "COPY.*TO PROGRAM" mail_logs
# Attachment flaw: Cisco's compromise check
cisco-esa> grep "System Security" system_logs
# On 15.5.5-014: engine must be 4.0.2 or later
cisco-esa> retroscannerstatus
# Must be 25.3.0.5525 or later
cisco-esa> contentscannerstatus
cisco-esa> contentscannerupdate

Cisco wants everyone below 16.5 on 16.5.0-780. Check the route before you book the window, because the two sets of notes differ. Cisco's October notes for 16.5.0-780 don't list any 15.5.5 build as a starting point, while its November notes for 16.5.0-825 list 15.5.5-014 and 16.0.4-302. If you're on 15.5.5-014, that makes 16.5.0-825 the documented route. Cisco also warns that some listed paths get pulled, so open the Deprovisioned AsyncOSdefinitionCisco's operating system for its Secure Email Gateway, on physical and virtual appliances alike. Check this version number first, since it decides whether you're fixed. Releases tab in its compatibility matrix first.

Follow this order. Upgrading first can wipe the evidence.

  1. 01
    Copy the evidence out

    Export mail_logs and system_logs from each gateway to somewhere the appliance can't touch. On a virtual gateway you suspect, capture forensic data before anything else, since a new instance destroys the old configuration and logs.

  2. 02
    Run the hunts and write down the results

    Do this before the upgrade, and do it on every cluster member. A clean search is only one data point, so go and read the firewall logs too.

  3. 03
    Update Content Scanner Tools

    Run contentscannerupdate if automatic updates are off. Cisco says this needs no software upgrade or restart, so there's no reason to wait for the change window.

  4. 04
    Save the configuration

    Keep a copy off the box. If encryption is enabled on the gateway, Cisco recommends keeping a saved copy, since encrypted values reset to factory defaults in some recovery situations. Don't load it onto a gateway you've rebuilt after a compromise, because Cisco says to rebuild without previous configuration files.

  5. 05
    Upgrade and reboot

    Go through System Administration, then System Upgrade, or run upgrade and DOWNLOADINSTALL in the CLI. Do one cluster member at a time and let each finish.

  6. 06
    Check the result

    Check the build and the Content Scanner Tools version, and make sure mail still flows. Then read through the GUI HTTPS cipher list, since Cisco says the upgrade rewrites its separators and wants someone to look.

chapter 04

Harden so the next one hurts less

BleepingComputer points out this is hardly the first time. Cisco patched another maximum-severity AsyncOS flaw in January, CVE-2025-20393, after attackers had been using it since November 2025. So assume the gateway will have another bad day, and limit what an attacker gets when it does.

before

Before: one gateway holds the keys

Management is reachable from the same interface as mail, the gateway's accounts are broad, and logs exist only on the appliance.

Before: one gateway holds the keys Management is reachable from the same interface as mail, the gateway's accounts are broad, and logs exist only on the appliance. 01 · INTERNET 02 · GATEWAY 03 · INSIDE root wipes traces stolen bind trusted relay Crafted email no login needed Secure Email Gateway mail and admin together Logs on the box attacker can edit Directory broad bind account Mail servers © Cohash · cohash.com
01 · INTERNET02 · GATEWAY03 · INSIDE Crafted email: no login needed Crafted email no login needed Secure Email Gateway: mail and admin together Secure EmailGateway mail and admintogether Logs on the box: attacker can edit Logs on the box attacker can edit Directory: broad bind account Directory broad bind account Mail servers Mail servers root stolen bind trusted relay © Cohash · cohash.com
after

After: patched, narrowed and watched

Mail and management are on separate interfaces, the directory account can only read what it needs, and logs leave the box as they're written.

After: patched, narrowed and watched Mail and management are on separate interfaces, the directory account can only read what it needs, and logs leave the box as they're written. 01 · INTERNET 02 · GATEWAY 03 · INSIDE fixed parser least privilege logs forwarded Inbound email Gateway on 16.5 admin interface split Directory read-only, rotated External log server out of reach © Cohash · cohash.com
01 · INTERNET02 · GATEWAY03 · INSIDE Inbound email Inbound email Gateway on 16.5: admin interface split Gateway on 16.5 admin interface split Directory: read-only, rotated Directory read-only, rotated External log server: out of reach External logserver out of reach fixed parser least privilege logs forwarded © Cohash · cohash.com

Durable fixes, each tied to this case

  1. 01
    Send logs off the appliance

    Cisco's advice to cross-check outside logs only works if those logs exist. Forward gateway logs to an external server, so someone with root on the gateway can't quietly edit the record.

  2. 02
    Split mail from management

    Put mail and admin on separate interfaces, and allow admin only from known, trusted hosts. Cisco's own hardening list asks for both.

  3. 03
    Shrink the gateway's accounts

    Make the LDAP bind account read-only and scoped to what lookups need. Put these logins and keys in a vault with a rotation schedule, so replacing them after an incident takes hours.

  4. 04
    Leave automatic updates on

    Content Scanner Tools fixes reach you without a software upgrade, and they only arrive by themselves if updates are enabled. Check that setting on every appliance.

  5. 05
    Stay on a line that gets fixes

    16.0 customers learned that a supported-looking line can stop getting builds for a given flaw. Keep a calendar of AsyncOS releases and budget one upgrade window a year.

chapter 05

If you've been breached

If you think a gateway was compromised

A hit in the logs, a System Security warning or strange traffic from the gateway all count. Patching won't remove someone who's already in, so work through these in order.

  1. 01Containfirst hour
    • Limit admin access to the gateway and block its outbound traffic, apart from what mail delivery and the investigation need.
    • Open a Cisco TAC case. If it's a physical appliance, turn on remote access so Cisco can get in and look.
    • Tell legal and your cyber insurer's contact early. Their clocks may already be running.
  2. 02Preservebefore any rebuild
    • On a virtual gateway, record forensic data first. Deploying a new instance destroys the configuration and logs.
    • Copy mail_logs and system_logs from every cluster member.
    • Save firewall, proxy and DNS logs covering the gateway, going back well before the Cisco disclosure date.
  3. 03Investigatefirst day
    • Don't trust the appliance's own logs. Root access lets an attacker remove or hide evidence.
    • Look for uploads from the gateway to outside addresses and downloads from malicious ones.
    • Check where the gateway's accounts logged in: directory, mail servers and anything using its SMTP login.
  4. 04Rebuildfirst days
    • Virtual: deploy a new VM on a fixed release and rebuild the configuration by hand. Cisco's file-write advisory says not to import old configuration files.
    • Physical: Cisco handles these through support, so do what TAC tells you. Don't improvise a rebuild.
    • Restore every member of the cluster, even if only one looks bad. Cluster SSH keys can be stolen and used on the others.
  5. 05Replace secretsthis week
    • Reissue certificates and private keys, and change the SAML certificate passphrase.
    • Change LDAP bind, RADIUS, SMTP authentication, SNMP, FTP Push and IPMI passwords.
    • Rotate DKIM signing keys and encryption keys, and create new Authentication API client credentials.
    • Reset local gateway admin passwords. Cisco's notes confirm local password hashes live on the appliance.
  6. 06Report and learnwithin your legal deadlines
    • Have counsel judge whether mail that passed through the gateway triggers regulator or customer notices.
    • Watch the rebuilt gateway for odd behavior, as Cisco advises, and write down what would have caught this sooner.
chapter 06

What this could look like: three typical setups

What this could look like: three typical setups

01 Crafted attachment 02 16.0.4-302 gateway old scanner tools 03 Files written asroot
scenario 01High risk
A mid-size firm on 16.0 that patched in September
the setup
A manufacturer's admin installed 16.0.4-302 the week the SQL bug came out and closed the ticket. File Reputation is on, and someone disabled automatic updates during an old change freeze.
what happens
Content Scanner Tools is still on an old version, so the attachment flaw stays open. A crafted attachment gets a file written to the gateway, and the September ticket says everything's done.
what to do
Run contentscannerupdate today and confirm 25.3.0.5525 or later. Search system_logs for the System Security message, read the firewall logs, then book the move to 16.5.
01 Member one log hit found 02 Cluster SSH keys 03 Save, then rebuild all members
scenario 02High risk
A hospital group with a virtual cluster
the setup
Three virtual gateways share a cluster. Someone greps mail_logs on one of them, and a line comes back that matches Cisco's example.
what happens
The team could rebuild only that gateway and move on. But cluster SSH keys could have been taken from it, so the other two members can't be trusted either.
what to do
Save forensic data first, then rebuild all three on a fixed release. Replace certificates, bind accounts and DKIM keys, and bring in counsel about patient data in mail.
01 Ciscoupgradescloud 02 Notice tocustomer 03 Old adminmailbox 04 Confirm inwriting
scenario 03Medium risk
A small firm on Cisco's hosted email security
the setup
A professional services firm uses Cisco Secure Email Cloud through a reseller. It has no command-line access and nobody there has thought about the gateway in years.
what happens
Cisco has already upgraded the service and has contacted customers where it found signs of compromise. If that notice went to a departed admin or an old shared mailbox, nobody read it.
what to do
Ask the reseller and Cisco, in writing, whether your tenant was flagged. Check who's listed as your admin contact. If you were flagged, renew the credentials and keys on the service.
chapter 07

How Cohash helps

We don't sell or run Cisco's email gateways, so your email team or provider does the upgrade. We come in afterwards and replace what the gateway knew. Our privileged access management team tracks down the bind accounts and service logins tied to appliances like this, moves them into a vault like CyberArk and schedules rotation. When the cleanup needs more hands, our CyberArk engineers can join for a few weeks. Not sure what your appliances can reach? Start with the free assessment.

key takeaways

What to take from this

  1. 01

    Both flaws are exploited, neither has a workaround, and either one can give an attacker root from a single email.

  2. 02

    16.0.4-302 covers only the SQL injection. Update Content Scanner Tools now and plan the move to 16.5.

  3. 03

    On 15.5.5-014, the version number isn't enough. Check that the Retro Remediation Engine is 4.0.2 or later.

  4. 04

    Hunt before you upgrade, and trust the firewall logs over the gateway's own.

  5. 05

    One compromised cluster member means restoring every member and replacing every key and password the gateway held.

platforms in this noteCyberArk →
fact-checked November 2026 against vendor and official documentation · where to go next
  1. How Cohash delivers CyberArk: scope, roles and architecture
  2. Hire vetted CyberArk engineers
  3. Privileged access management
  4. Book a 30-minute working session with a senior engineer
01Who runs our email filter, and how can I tell if it's Cisco?

Ask whoever runs your email. You can also look up your domain's MX records, which show where your mail goes first, though a hosted service may hide the product name. Older paperwork might call it an IronPort appliance, since that was the product's earlier name.

02Does one of our people have to click on something first?

No. The gateway reads the message as part of handling it, so the attack can work before the email reaches anyone's inbox. Training staff to spot phishing won't help with this one.

03We're already on 16.0.4-302. Doesn't that cover us?

No. That build fixes the SQL injection. For the attachment flaw, Cisco has no 16.0 build, so you need Content Scanner Tools 25.3.0.5525 or later if File Reputation is on. Cisco's hardening release also tells 16.0 customers to migrate, so plan on 16.5.

04We're on 15.5.5-014. Why would we need to check again?

Cisco updated the fix inside that release on October 31. Run retroscannerstatus. If the Retro Remediation Engine is 4.0.2 or later, you have the update. Cisco still recommends moving to 16.5, where the engine version doesn't matter.

05Cisco hosts our email security. Do we need to do anything?

Cisco says it has moved its cloud devices to 16.5.0-780, so patching isn't on you. You do need to find out whether Cisco contacted you about malicious activity. If it did, renew the credentials and keys on the service.

06My logs are clean. Am I safe?

Clean logs are good news, but they aren't proof. Cisco says an attacker with root can remove or hide evidence, and its search pattern is only an example. Check firewall and network logs from outside the gateway before you sign off.

07Is Secure Email and Web Manager affected?

Cisco says neither exploited flaw touches Secure Email and Web Manager or Secure Web Appliance. The hardening release is different: it does cover Secure Email and Web Manager, with fixes in 15.5.5-006 and 16.5.0-429.

next track

That's the end of this track.

Next up: running enterprise platforms well. ERP, CRM and ITSM work that holds up after go-live.

  1. 01HashiCorp Vault 2.2 explained: TPM login, post-quantum PKI and the October fixes
  2. 02Upgrading to SailPoint IdentityIQ 9.0: what breaks and what to fix first
  3. 03Terraform 1.17 explained: what's new, who it touches and how we'd upgrade
  4. 04BeyondTrust Remote Support zero-day: patch it, then check who got in
  5. 05Patch PHP and Node.js this week: what broke and who's exposed
  6. 06OpenShift support dates through 2027: which clusters run out first and how to plan the upgrades
  7. 07Oracle's October security patches: what to fix first in EBS, PeopleSoft and Access Manager
  8. 08Oracle Java drops four root certificates over eMudhra: patch, then test what breaks
  9. 09Delinea Secret Server auth bypass: why the FIDO2 hotfix isn't enough
  10. 10IBM Verify Access flaws let attackers run code without a login: patch to 11.0.3.2
  11. 11Foundry Workflows retire on December 1. We'd start moving yours to Agent Framework now
  12. 12Vertex AI is retiring text-embedding-005 and three other embedding models: how to re-embed before search breaks
  13. 13dbt drops v1.3 to v1.7 from its platform on January 31, 2027: get pinned jobs onto a release track before they're deleted
  14. 14Brotli CVE-2026-111497 is now on CISA's exploited list, and you probably run more copies than you think
  15. 15Auth0 moved the Rules and Hooks deadline: Hooks stop on 30 April 2027, Rules on 31 January 2028
  16. 16Still on Okta Classic Engine? Three features stop working on March 5, 2027, and two of them block the upgrade
  17. 17Patched GitLab, still exposed: the CI regex bug CISA lists as exploited, and the Ruby step people miss
  18. 18Patched Ivanti Connect Secure in October? Two more exploited bugs, and the AD cleanup
  19. 19Patch roundup, week of Oct 26: upgrade Secret Server today, then work down the list
  20. 20UiPath 2023.10 support has run out and 2024.10 is next. What we'd upgrade first
  21. 21Password Safe 24.3 runs out of support in December, and 26.3 is still one upgrade away
  22. 22Upgrade Entra Connect Sync before April 2027 and move users off Microsoft's SMS codes
  23. 23Chrome drops XSLT on 17 November. What breaks in InterSystems, and the fix for your version
  24. 24Agent Builder, Evals and saved prompts shut on 30 November. Here's what to pull out first
  25. 25BeyondTrust Remote Support and PRA 24.3 lose support in November, and there's no one-step upgrade
  26. 26OpenShift 4.19 is running out of support. Do you park on 4.22 or keep going to 5.0?
  27. 27Terraform Enterprise 2.1 moves VCS ingress into your agents. What to change before October 2027
  28. 28Mule 4.12 Edge leaves standard support this month. Upgrade to 4.13 before CloudHub does it for you
  29. 29RSA SecurID primary support ends in stages from December. Here's how you get to 8.9
  30. 30Delinea Secret Server login bypass: why 12.2.7 won't save you from this 9.9
  31. 31Entra password reset won't take directory phone numbers from February 2027. Who gets stuck?
  32. 32Linux kernel CVE-2026-98432: why switching off MPTCP won't save you, and what to patch
  33. 33Patch roundup, week of Nov 2: fix Linux and WS_FTP first, then work down the list
  34. 34Auth0's lazy migration changes on 20 January 2027, and half-imported users won't fix themselves
  35. 35Your Databricks AI Functions need Runtime 17.3 LTS by January 2027, even if 15.4 is still supported
  36. 36SonicWall's SSL VPN bug lets a valid password skip SAML and your MFA
  37. 37You patched your Cisco email gateway. One email can still take it over.
all learning tracks →
Time to upgrade or move No newenvironments Suspended Decommission… Nov 2026 Oct 2027
Engineering · 15 minDataverse for Teams is going away: what to keep, move or let go before September 2027

Most teams still have time to upgrade or move their apps. After September they don't.

read next
next step

Tell us what you're running. We'll tell you what we'd do first.

A 30-minute working session with a senior engineer. No slides.

useful to bring
  • The platform and the version you're on
  • Where you need to get to, and by when
  • What's custom: rules, plugins, integrations
  • Any audit finding or deadline driving it