Drop pre-checks in stratis-cli for opposite tier and same tier device adds

[mulkieran]

The stratisd error messages look like this without the expensive pre-check:

# PYTHONPATH=./src ./bin/stratis blockdev list
Pool Name   Device Node   Physical Size    Tier   UUID                                
p0          /dev/sda5        865.33 GiB    DATA   91f548a6-cb57-4aac-8bb3-14579bd82005
pn          /dev/sda2         15.70 GiB    DATA   f525d22c-d1e4-43b7-8c4f-c06891ce0378
pn          /dev/sda3            50 GiB   CACHE   08eb0411-3795-4ec9-8bc6-c9f21cecea72

# PYTHONPATH=./src ./bin/stratis pool add-data p0 /dev/sda3
Execution failed:
stratisd failed to perform the operation that you requested. It returned the following information via the D-Bus: ERROR: Some devices specified appear to be already in use by other Stratis pools: devices (/dev/sda3) appear to belong to Stratis pool with UUID dccccb6e-e82b-4f7d-b92b-21ccc2ff8aea.

# PYTHONPATH=./src ./bin/stratis pool add-cache pn /dev/sda5
Execution failed:
stratisd failed to perform the operation that you requested. It returned the following information via the D-Bus: ERROR: Some devices specified appear to be already in use by other Stratis pools: devices (/dev/sda5) appear to belong to Stratis pool with UUID 972f8338-8f60-4394-9a0f-b360e7e63853.

# PYTHONPATH=./src ./bin/stratis pool add-cache pn /dev/sda2
Execution failed:
stratisd failed to perform the operation that you requested. It returned the following information via the D-Bus: ERROR: Devices (/dev/sda2) appear to be already in use by this pool which has UUID dccccb6e-e82b-4f7d-b92b-21ccc2ff8aea, but in the data tier not the cache tier.

# PYTHONPATH=./src ./bin/stratis pool add-data pn /dev/sda3
Execution failed:
stratisd failed to perform the operation that you requested. It returned the following information via the D-Bus: ERROR: Devices (/dev/sda3) appear to be already in use by this pool which has UUID dccccb6e-e82b-4f7d-b92b-21ccc2ff8aea, but in the cache tier not the data tier.

# PYTHONPATH=./src ./bin/stratis pool add-data pn /dev/sda5
Execution failed:
stratisd failed to perform the operation that you requested. It returned the following information via the D-Bus: ERROR: Some devices specified appear to be already in use by other Stratis pools: devices (/dev/sda5) appear to belong to Stratis pool with UUID 972f8338-8f60-4394-9a0f-b360e7e63853.

# PYTHONPATH=./src ./bin/stratis pool add-data p0 /dev/sda5
Execution failed:
stratisd reported that it did not execute every action that it would have been expected to execute as a result of the command that you requested: Expected to add the specified blockdevs to the data tier in pool p0 but stratisd reports that it did not actually add some or all of the blockdevs requested; devices added: (), devices requested: (/dev/sda5)

This is how it looks now, with the expensive pre-check:

# PYTHONPATH=./src ./bin/stratis pool add-data p0 /dev/sda3
Execution failed:
It appears that you issued an unintended command: At least one of the provided devices is already owned by an existing pool's CACHE tier; device /dev/sda3 would be added to the DATA tier but is already in use in the CACHE tier of pool pn

# PYTHONPATH=./src ./bin/stratis pool add-cache pn /dev/sda5
Execution failed:
It appears that you issued an unintended command: At least one of the provided devices is already owned by an existing pool's DATA tier; device /dev/sda5 would be added to the CACHE tier but is already in use in the DATA tier of pool p0

 PYTHONPATH=./src ./bin/stratis pool add-cache pn /dev/sda2
Execution failed:
It appears that you issued an unintended command: At least one of the provided devices is already owned by an existing pool's DATA tier; device /dev/sda2 would be added to the CACHE tier but is already in use in the DATA tier of pool pn

# PYTHONPATH=./src ./bin/stratis pool add-data pn /dev/sda3
Execution failed:
It appears that you issued an unintended command: At least one of the provided devices is already owned by an existing pool's CACHE tier; device /dev/sda3 would be added to the DATA tier but is already in use in the CACHE tier of pool pn

# PYTHONPATH=./src ./bin/stratis pool add-data pn /dev/sda3
Execution failed:
It appears that you issued an unintended command: At least one of the provided devices is already owned by an existing pool's CACHE tier; device /dev/sda3 would be added to the DATA tier but is already in use in the CACHE tier of pool pn

# PYTHONPATH=./src ./bin/stratis pool add-data pn /dev/sda5
Execution failed:
It appears that you issued an unintended command: At least one of the provided devices is already owned by an existing pool's DATA tier; device /dev/sda5 would be added to the DATA tier but is already in use in the DATA tier of pool p0

The one substantive difference is that the pool name is available in the stratis-cli side error message but not the stratisd error message, which uses the UUID. It's possible the stratisd error message could give the device UUID, which might be helpful.

[coderabbitai]

📝 CodeRabbit Plan Mode

Generate an implementation plan and prompts that you can use with your favorite coding agent.

• Create Plan

Examples

• Example 1
• Example 2

---

🔗 Similar Issues

Related Issues

• Unable to Start Pool After Cache Device Lost – No Recovery or Force Option Available #767
• Low-level devicemapper error returned when creating filesystem with size that is divisible by 512 but not by 4096 #760

👤 Suggested Assignees

• rismalrv
• mulkieran

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!

[mulkieran]

The pre-checks are paid every time and are almost always unnecessary. Post-checks in error-reporting would do just as well.

[mulkieran]

We could also drop the part where we raise a StratisCliNameConflictError, because the IncoherenceError, which we could rename to something else, or collapse into the NoChangeError, would do just as well.