Back to Resources
Blog

MEM Is Gone. Oracle's MySQL Plug-in Is Next. The Capabilities Don't Have to Be.

Andre Beaumont·July 16, 2026

A blueprint-style illustration titled "Picking Up Where the Plug-in Leaves Off": a hand-drawn suspension bridge whose near half is plain line-art and whose far half glows bright cyan, carrying an Integration Plumbers panel across to an Oracle Enterprise Manager structure, captioned that the MySQL Plug-in for Oracle Enterprise Manager from Integration Plumbers enters Early Access July 31, 2026

If you ran MySQL Enterprise Monitor, you've made this move once already. MEM reached end of life in January 2025, and Oracle's guidance was explicit: migrate to Oracle Enterprise Manager for MySQL, the plug-in that brought your MySQL targets into the OEM console. A lot of teams did exactly that. It was the sensible call, since the plug-in came at no extra cost with a commercial edition and ran in the console they already operated for the rest of the estate.

Now the EM plug-in is being deprecated as well. You won't find that on a public Oracle page yet; the notice went out through My Oracle Support, where customers with a support contract can see it. If you run the plug-in and this is news, check MOS or ask your Oracle account team. That's two MySQL monitoring products wound down in eighteen months, and the second one was the official replacement for the first.

Here's the part that matters if you're the one running these systems. The MySQL Plug-in for Oracle Enterprise Manager from Integration Plumbers, entering Early Access on July 31, 2026, picks up where Oracle's plug-in leaves off. It runs in the same console with the same operational model, restores the MEM capabilities the Oracle plug-in never carried (backup visibility, curated thresholds, version end-of-life advisories), and its collection layer is built for MySQL 8.4, where monitoring written for older releases breaks quietly.

This post has four parts: how we got from MEM to here, what the Oracle plug-in covered and what it never did, the 8.4 change that breaks old monitoring, and what our plug-in restores and builds forward.



1. How we got here

MySQL Enterprise Monitor was the closest thing the MySQL world had to a purpose-built monitoring console. It shipped sensible alert thresholds out of the box, it watched your backups, and it ran a catalog of advisors that told you when an instance was misconfigured or running on a version that had aged out of support. It reached end of life on January 1, 2025, and the documentation that used to live on dev.mysql.com followed it out the door in early 2026; the pages survive now mostly in the downloadable manual and the Wayback Machine.

Oracle's EOL notice told MEM users where to go next: Oracle Enterprise Manager for MySQL, the plug-in that puts MySQL databases, InnoDB Clusters, and ClusterSets into the OEM console as first-class targets. For enterprises already running OEM for the rest of their estate, this was a reasonable landing spot, and it was actively maintained. The 13.5.6 release shipped in April 2026, the same month Oracle revised the plug-in's user guide for the EM 24ai platform.

Then came the deprecation notice, delivered through My Oracle Support rather than announced publicly. Teams that made the move in good faith after MEM's retirement are being asked to plan another one, and this time there is no like-for-like replacement inside the EM console to point to.

2. What the bridge carried, and what it never did

A blueprint-style capability matrix comparing MySQL Enterprise Monitor (marked EOL January 2025), the Oracle EM plug-in, and Integration Plumbers across five rows: OEM integration, backup dashboard (backup_history), version EOL advisor, MySQL 8.4 native syntax, and least-privilege security footprint. The Oracle plug-in shows a red X on the backup dashboard, the version EOL advisor, 8.4 native syntax, and the security footprint, while the Integration Plumbers column shows a check on every row

Credit where due: the Oracle plug-in was a real product. It collected around 500 metrics across roughly ninety categories, shipped a Query Analyzer, monitored InnoDB Cluster and ClusterSet topologies, supported autodiscovery, and could watch a remote MySQL host without installing an agent on it.

But it was never the whole of MEM, and Oracle's own documentation says so. The user guide describes the plug-in as based on MySQL Enterprise Monitor but containing "only a subset of MEM Advisors and graphs." The subset had a shape, and if you came from MEM, the missing pieces were exactly the ones you noticed:

  • The backup dashboard didn't come across. MEM read the backup_history and backup_progress tables that MySQL Enterprise Backup maintains and turned them into a dashboard: last successful full, last failure, run time and lock time history, and a coverage view that flagged instances with no configured backup at all. The Oracle plug-in has no equivalent. You get alerted on CPU and connections, not on "no successful backup in four days."
  • The advisor catalog was cut down hard. MEM ran fourteen advisor categories. The plug-in ships five compliance rule groups totaling twenty-nine rules. Security alone went from roughly two dozen MEM advisors (anonymous accounts, root without password, insecure password hashes, privilege alterations) to five rules.
  • The version advisor didn't make the cut. MEM had an advisor that flagged when an instance was no longer eligible for Oracle Premier Support because its version had aged out. With 8.0 community support ending in April 2026, that is precisely the check estates needed this year, and the plug-in doesn't have it.

So even before the deprecation notice, teams that moved from MEM to the plug-in were living with a gap. The deprecation just sets a clock on the parts that did carry over.

3. The 8.4 change that breaks old monitoring

A blueprint-style diagram titled "Architectural Advantage: Built 8.4-First": a toggle switch marked ENGAGED routing a bright 8.4 Native Layer line to the current SHOW REPLICA STATUS and SHOW BINARY LOG STATUS commands, while a dashed 8.0 Deprecated Layer runs below as a version-gated path, captioned "A clean rebuild, not a patched legacy tool. Keeps working seamlessly through the version you are migrating onto"

While all of this was happening, MySQL itself moved. 8.0 community support ended in April 2026, 8.4 is the LTS release everyone is migrating onto (supported through April 2029), and 8.4 did something monitoring tools have to reckon with: it removed the legacy replication interface.

SHOW SLAVE STATUS is gone. So are the Com_slave_* status counters and the master_info_repository settings that a lot of monitoring was built on. The replacement syntax (SHOW REPLICA STATUS, SHOW BINARY LOG STATUS, the REPLICA/SOURCE terminology) has been available for years, but plenty of collection code never moved to it.

What breaks depends on how your monitoring was written, and neither way is good. Tooling that still issues the removed commands gets a syntax error on 8.4: the query fails outright, which at least announces itself. Tooling that reads the old Com_slave_* counters gets something quieter and worse, because those variables simply stop appearing in SHOW STATUS. The metric doesn't error. It goes blank, which reads as "fine" right up until you need it during a failover.

This is the real cost of monitoring a moving database with a deprecated tool. A product in deprecation stops tracking the server's changes, and a version migration is exactly when you most want your monitoring to be honest.

4. Picking up where the plug-in leaves off

A blueprint-style diagram titled "Day-One Value with Curated Thresholds and Version Tracking": a left panel of pre-set curated defaults (Replication Lag flagged CRITICAL, Connection Saturation and Buffer-Pool Health flagged WARNING) beside a right-panel Version End-of-Life Advisor listing Instance A on 8.4 LTS and Instance B on 8.2 LTS as healthy and Instance C on 8.0 Community flagged EOL EXPIRED, annotated "Hand this straight to an auditor"

The MySQL Plug-in for Oracle Enterprise Manager from Integration Plumbers is built for the operators this history has been happening to. If you deployed Oracle's plug-in, the model will feel familiar: MySQL targets in the OEM console you already run, monitored through your existing agents, feeding the incident rules and notification chains you already maintain. No new console to learn, no separate pane of glass to justify.

On that foundation, the capabilities MEM was actually valued for come back:

  • Your backups are visible again. The plug-in watches MySQL Enterprise Backup (and Percona XtraBackup) history and alerts on the things that matter: a failed backup, no successful backup in N days, a cluster member that isn't being backed up at all. This is the piece neither the Oracle plug-in nor any major APM tool picked up when MEM retired.
  • The thresholds ship set. Curated WARNING and CRITICAL defaults arrive with the plug-in for replication lag, connection saturation, buffer-pool health, and backup age, so a freshly added target is useful on day one instead of silent until you tune it.
  • End-of-life becomes a screen. A version advisory flags instances running on a release that's aged out of support, the same idea as MEM's "no longer eligible" advisor, now a check you can hand to an auditor. The fuller advisor catalog that made MEM valuable is being rebuilt on OEM's own compliance framework as the next layer on top.

And it's a clean rebuild, which is an advantage. The collection layer is built 8.4-first, speaking the current replication syntax so it keeps working through the version you're migrating onto, with the older 8.0 tier handled as a deprecated, version-gated path rather than an afterthought. It runs from a single read-only grant (SELECT, PROCESS, REPLICATION CLIENT), a smaller footprint than the five privileges Oracle's plug-in required, with no SUPER, TLS that fails closed rather than silently dropping to plaintext, and your choice of a local or remote agent.


MEM is not coming back, and the plug-in Oracle offered in its place is now on the same road. But the reasons you ran them don't have to retire with the products: knowing your backups are good, getting alerted before a threshold becomes an outage, being told when an instance has aged off a supported version. Those capabilities stay in the console you already use for the rest of your estate, rebuilt for the MySQL version you're actually running, on a roadmap that gets longer from here instead of shorter.

If you've re-platformed your MySQL monitoring once since 2025, you shouldn't have to keep doing it. That's the point of building this inside Enterprise Manager rather than beside it.

This is the fourth post in a series on the blind spots your estate monitoring leaves across MySQL, SQL Server, and DB2. The earlier three were The Backup You Can't See, on whether last night's backup actually succeeded; The DR-Readiness Gap, on whether a "synchronized" cluster could really fail over without losing data; and SQL Server 2016 End of Support, on the July 14 version cliff. This one traced the path from MEM's retirement through the Oracle plug-in's deprecation, what that plug-in carried and what it never did, the 8.4 change that breaks naive monitoring, and what our plug-in brings back into Oracle Enterprise Manager.

Get Early Access

Backup monitoring, curated thresholds, the version-EOL advisory, and an 8.4-safe collection layer arrive with the MySQL Plug-in for Oracle Enterprise Manager, and Early Access opens July 31, 2026. Re-platformed once already? Get this one ready before you have to move again. Learn more and sign up:

We bring the same least-privilege, TLS-first approach to the Microsoft SQL Server and DB2 plug-ins, too.

Next in the series, no cliff, no drama: monitoring DB2 11.5 and 12.1, and the new ground ahead in z/OS and AI Query Optimizer coverage.


Didn't Oracle already replace MEM? Is that replacement being discontinued too?+

Yes to both, which is the point of this post. When MySQL Enterprise Monitor reached end of life in January 2025, Oracle directed customers to Oracle Enterprise Manager for MySQL, its EM plug-in, and many teams migrated. A deprecation notice has now landed for that plug-in as well, delivered through My Oracle Support rather than a public Oracle page. If you run the plug-in, confirm the notice through MOS or your Oracle account team.

What did Oracle's plug-in not carry over from MEM?+

Oracle's own user guide describes the plug-in as containing "only a subset of MEM Advisors and graphs." In practice the gaps were the ones MEM users relied on most: there's no backup dashboard, the advisor catalog was cut from fourteen categories to five compliance rule groups (twenty-nine rules total, with Security dropping from roughly two dozen advisors to five), and the version-EOL advisor that flagged unsupported instances didn't make the cut.

Why does MySQL 8.4 break monitoring that worked on 8.0?+

8.4 removed the legacy replication interface. SHOW SLAVE STATUS, the Com_slave_* status counters, and the master_info_repository settings are gone. Collection code that still issues the old commands gets a syntax error; code that reads the old counters gets something worse, because those variables simply stop appearing in SHOW STATUS, so the metric goes blank instead of erroring. A tool that's in deprecation won't be updated to follow those changes.

Does your plug-in monitor backups the way MEM did?+

Yes, and it's the capability neither the Oracle plug-in nor a general APM tool picked up when MEM retired. It reads MySQL Enterprise Backup history (and Percona XtraBackup) and alerts on what matters: a failed backup, no successful backup in N days, or a cluster member that isn't being backed up at all, collected over a standard connection with no agent on the database host.

I'm on Oracle's plug-in today. Why move now instead of later?+

Two reasons stack up. The deprecation notice puts a clock on the plug-in, and it was already missing MEM capabilities you may be working around. Meanwhile 8.4 is changing the replication surface underneath any tool that isn't actively tracking it. Moving to a plug-in that's built 8.4-first and restores the backup, threshold, and version-EOL capabilities means you re-platform once, deliberately, rather than again under pressure during a migration.

What access does the plug-in need on each instance?+

One read-only login with SELECT, PROCESS, REPLICATION CLIENT, a smaller footprint than the five privileges Oracle's plug-in required. There's no SUPER and no operating-system access, and the TLS connection fails closed rather than silently dropping to plaintext. You can point it at each instance through a local or a remote agent, whichever fits your topology.

Not Sure Where to Start?

Take our free OTEL Maturity Assessment to identify gaps and get a personalized action plan.

Take the Free Assessment