-
Notifications
You must be signed in to change notification settings - Fork 14
Labels
Block NodeIssues/PR related to the Block Node.Issues/PR related to the Block Node.Block ProviderIssue related to Block ProvidersIssue related to Block ProvidersImprovementCode changes driven by non business requirementsCode changes driven by non business requirements
Description
Story Form
As a Block Node Developer/User/Operator
I want to have reliable BlockAccessors
So that I can trust the reliability of my system
Technical Notes
- see this comment thread
- currently
BlockAccessorsare generally created only when the actual block they target exists- this is correct, but that does not handle the case where during the lifetime of the accessor, the actual block may be rendered unaccessible (the reason does not matter, deleted, moved, filesystem issue etc.)
- in those cases we need to be absolutely sure that the system will not fail as this is a very much expected condition
- we must agree on API level what should be done in such cases
- for now, what makes sense is that accessors never throw any exceptions, except for bugs and unrecoverable cases In such cases, the node itself cannot recover?
- we should return
nullwhen we have an accessor, but attempts to read the bytes are not successful
Metadata
Metadata
Assignees
Labels
Block NodeIssues/PR related to the Block Node.Issues/PR related to the Block Node.Block ProviderIssue related to Block ProvidersIssue related to Block ProvidersImprovementCode changes driven by non business requirementsCode changes driven by non business requirements