business

Verdict

Submitted 6/8/2026, 8:04:11 AM · Completed 6/8/2026, 8:05:39 AM

5.5
pivot
The idea

What Could Cause an InnoDB .frm and .ibd File Mismatch

Pain point
InnoDB .frm and .ibd file mismatches can lead to data inconsistency and loss without error logs indicating the issue.
Who has this problem
Database administrators managing MariaDB with innodb_file_per_table enabled
Contradiction (TRIZ)
Need for consistent backups without requiring full table scans or error logs to detect corruption
Ideal final result
A system that automatically detects and resolves .frm and .ibd file mismatches in real-time without disrupting database operations
Suggested solution
Implement a scheduled script that periodically checks .frm and .ibd file consistency using checksums and database metadata, and automatically repairs or alerts on discrepancies.
Show original source text →
On June 3rd we noticed that one of our tables was not showing up in SHOW TABLES, but a direct SELECT * FROM table_name returned data normally. When we ran FLUSH TABLES the select command stopped working. (Our interpretation is that the table handle was still open in the InnoDB buffer) Then we inspected /var/lib/mysql/<database_name>/ and found that a significant number of tables across the database were in inconsistent states: Some tables had both .frm and .ibd → worked fine Some tables had only .frm , no .ibd → visible in SHOW TABLES , failed on query Some tables had only .ibd , no .frm → not visible in SHOW TABLES , but still queryable until FLUSH TABLES What the error log tells us: The server had been running continuously since March, the first hard evidence of the problem appears at 2026-06-01 at 02:28 AM: 2026-06-01 02:28:35 [ERROR] InnoDB: Cannot rename './database_name/xxxxxxxxxxx.ibd' to './database_name/#sql-ib867802.ibd' because the source file does not exist. 2026-06-01 02:28:35 [ERROR] InnoDB: Cannot rename './database_name/xxxxxxxx.ibd' to './database_name/#sql-ib867803.ibd' because the source file does not exist. [... 15 more tables with the same error ...] However, we did not find any evidence of errors before that. That is, we noticed it on June 3rd, but there is evidence that it was already corrupted on June 1st. Then, on the 3rd, we restarted the server and the InnoDB startup scan confirmed the damage: 2026-06-03 10:17:53 [ERROR] InnoDB: Cannot open datafile for read-only: './database_name/xxxxxxxxxxxxxxxxxx.ibd' OS error: 71 2026-06-03 10:17:53 [ERROR] InnoDB: Could not find a valid tablespace file for database_name/xxxxxxxxxxxxxxxx. Additionally, after reviewing the backup sizes, we found that between May 28th and May 29th the backup size decreased by approximately 200 MB: 2026-05-24 7.724 GiB 2026-05-25 7.724 GiB 2026-05-26 7.725 GiB 2026-05-27 7.727 GiB 2026-05-28 7.729 GiB 2026-05-29 7.530 GiB 2026-05-30 7.531 GiB 2026-05-31 7.532 GiB Questions Finally, we were able to recover a backup from May 28th with the complete information, loosing some data. Bc mysqldump silently skipped the affected tables. Tables that only had an .ibd file (with no corresponding .frm file) were completely invisible to mysqldump : they did not appear in SHOW TABLES or in information_schema --> so no error was triggered 😭 What non-crash scenarios can cause .ibd files to disappear silently while MariaDB is running, with no errors in the error log? The disk was not full. The process ran continuously for ~3 months. No kernel panics or OOM events. The files simply vanished for a subset of tables at some point between the end of the May 28th backup and June 1st. Do people verify .frm / .ibd consistency before backups are taken in production environments? We always assumed that a successful mysqldump meant the backup was valid and complete. We now know that this is not necessarily the case :( Environment: OS: Debian 12 MariaDB: 10.11.14-MariaDB-0+deb12u2-log innodb_file_per_table = ON log_bin = OFF Server role: standalone (no replication)
TRIZ inventive level: 3/5· Principles: parameter changes, self-service
Synthesis verdict
**Pivot**: The idea of building a tool to verify .frm and .ibd consistency has potential, but the current proposal lacks a clear business venture, revenue model, and market opportunity. The problem described is real and painful for database administrators, but it does not translate into a viable business venture unless framed as a tool, service, or automation to prevent or detect such corruption. A solo or 2-person team can build a tool to verify .frm and .ibd consistency within 4-12 weeks, but integrating this check into a production environment and ensuring it doesn't interfere with normal database operations could be challenging.

Strengths

  • A solo or 2-person team can build a tool to verify .frm and .ibd consistency within 4-12 weeks
  • The problem targets a rare but critical failure mode—silent disappearance of InnoDB .ibd files that bypasses existing error logs and backup validation
  • A dedicated daemon that periodically cross-checks the directory listing against information_schema.tables (or performs checksum-based validation) and raises immediate alerts, optionally auto-triggering recovery procedures, fills a clear gap

Weaknesses

  • The idea describes a technical problem rather than a monetizable business venture
  • There is no revenue model, pricing strategy, or value-capture path
  • The venture faces existential risk due to undefined scope and potential regulatory, platform, and market challenges

Best angle

The venture should focus on building an automated, proactive InnoDB metadata integrity validator for self-managed MySQL/MariaDB environments, with a clear revenue model and market opportunity.

Panel verdicts

Competition

nvidia/nemotron-3-nano-omni-30b-a3b-reasoning

7.0

A real‑time, file‑level consistency monitor that validates .ibd existence against InnoDB metadata provides a unique, durable differentiation beyond generic database monitoring tools.

The problem targets a rare but critical failure mode—silent disappearance of InnoDB .ibd files that bypasses existing error logs and backup validation. Current solutions are generic monitoring platforms (e.g., Datadog, Prometheus exporters) or backup‑verification tools (Percona XtraBackup, mysqldump) that do not continuously compare filesystem inode presence with InnoDB metadata. A dedicated daemon that periodically cross‑checks the directory listing against information_schema.tables (or performs checksum‑based validation) and raises immediate alerts, optionally auto‑triggering recovery procedures, fills a clear gap. Competitors such as Percona Monitoring and Management, Datadog’s MySQL integration, and traditional backup suites lack this specific file‑level consistency check, giving the entrant a defensible differentiation. The niche focus on .ibd disappearance makes the market small but the pain point high, supporting durability if the product is reliable and integrates with existing backup pipelines. However, widespread adoption depends on proving low false‑positive rates and offering easy deployment, so the differentiation is real but may face limited scalability.

Market

qwen/qwen3-next-80b-a3b-instruct

2.0

The real opportunity lies not in describing the corruption, but in building an automated, proactive InnoDB metadata integrity validator for self-managed MySQL/MariaDB environments — but that’s not what’s being proposed here.

This is not a business idea — it's a post-mortem analysis of a catastrophic database corruption incident. The user is describing a technical failure mode in MariaDB/InnoDB, not proposing a product, service, or market solution. There is no target audience, no customer need being addressed, and no monetizable offering. While the problem described (silent .ibd file loss, inconsistent metadata, mysqldump blind spots) is real and painful for database administrators, it does not translate into a viable business venture unless framed as a tool, service, or automation to prevent or detect such corruption. As-is, this is an internal operational lesson, not a market opportunity. Even if one were to build a product around this — e.g., a MariaDB integrity checker that validates .frm/.ibd consistency pre-backup — the market is extremely narrow: only large, self-managed MySQL/MariaDB deployments with no replication or HA would care, and even then, they’d likely build this internally. The total addressable market is likely under 10,000 organizations globally, with low willingness to pay for a niche diagnostic tool. The emotional weight of the story underscores a real pain point, but the idea as presented lacks any product, pricing, or customer segment.

Risk

nvidia/llama-3.3-nemotron-super-49b-v1

2.0

Undefined venture scope coupled with potential regulatory, platform, and market challenges threatens viability.

The proposed business venture is based on a specific, resolved MySQL database corruption incident. However, the idea lacks clarity on what the actual venture is (e.g., consulting, software development for prevention/detection tools, training). Without a defined product/service, the venture faces existential risk. Assuming the venture aims to develop tools/services around database integrity, **key failure modes** could be: 1. **Regulation**: Emerging or tightened data protection regulations (e.g., GDPR, CCPA) might require solutions that are more comprehensive than what the venture can initially offer, leading to non-compliance. 2. **Platform Risk**: Dependence on MySQL/MariaDB ecosystems; if these databases lose market share to cloud-native or NoSQL solutions, the venture's relevance diminishes. 3. **No-Budget Customers**: Target market (likely small to medium businesses) might not allocate budget for preventive database integrity tools, prioritizing recovery services instead.

Monetization

mistralai/mistral-medium-3.5-128b

3.0

A technical problem alone does not constitute a business venture without a monetizable solution.

The idea describes a technical problem (inconsistent MySQL table states) rather than a monetizable business venture. There is no revenue model, pricing strategy, or value-capture path. The scenario lacks a product, service, or solution to sell—only a post-mortem analysis of a database corruption issue. Without a defined offering (e.g., a tool to detect .frm/.ibd inconsistencies, a backup validation service, or consulting), there is no basis for monetization. Even if framed as a potential SaaS (e.g., 'database health monitoring'), the description provides no concrete pricing, channels, or unit economics. The value here is purely diagnostic, not commercial.

Viability

meta/llama-4-maverick-17b-128e-instruct

8.0

A solo or 2-person team can build a tool to verify .frm and .ibd consistency within 4-12 weeks by leveraging existing utilities and focusing on integration and testing.

The task is to identify non-crash scenarios that can cause .ibd files to disappear silently while MariaDB is running. The given context suggests that the issue occurred between May 28th and June 1st without any errors in the error log. To build a tool or solution that detects such inconsistencies, a solo or 2-person team would need to develop a script or utility that checks for .frm and .ibd consistency. This task is feasible within 4-12 weeks. The team can leverage existing open-source tools or MariaDB's own utilities as a starting point. However, integrating this check into a production environment and ensuring it doesn't interfere with normal database operations could be challenging. The team would need to consider factors like performance impact, handling edge cases, and testing thoroughly. Despite these challenges, the task is still achievable with a focused effort.

Synthesized by meta/llama-3.3-70b-instruct · 44.6s