1. Main development threads
UNDER CONSIDERATION | PLANNED | STARTED | FINISHED |
This page presents a status of the ongoing developments
Status | Priority | Summary | JIRA Ticket |
---|---|---|---|
HIGH | Refactoring the Configuration service | ||
HIGH | Redesign of Qserv's "Empty Chunk Lists" into "Chunk Inventory" | ||
MEDIUM | Add chunk versioning | Create a ticket | |
CRITICAL | Huge memory leak in the Master Replication Controller when Qserv workers are down | ||
HIGH | Integration test for the Replication System | Create a ticket | |
MEDIUM | Security reinforced protocol for Controllers/Workers communications and worker database operations | ||
Reinforce the controllers / workers and file client / file server protocols to carry over and verify a unique identifier of a Qserv instance, so that the Replication system's configuration errors were properly detected and handled, and cross-talks between two or many instances of the Replication system were prevented. | |||
A Design of the Qserv Monitoring System | |||
Implement Qserv Monitoring Services (the Dashboard) | |||
| |||
Master replication controller with built-in REST API support | |||
Deploying, configuring and operating the Qserv Replication system | |||
Improved container build procedure for the Replication System's tools | |||
Migrate binary tools of the Replication system to the common Application framework | |||
Add REST services for monitoring and managing the Replication System | |||
Minor code refactoring | |||
Evaluate MySQL table partitioning for the reliable ingest with the rollback option | No ticket required for this effort. All details are found within a document linked on this row. | ||
Worker database management | |||
Implement a framework for sampling monitoring parameters of Qserv workers | |||
Add support for super-transactions | |||
Integrate Qserv replication system with CSS | |||
Log REST requests to the Master Replication Controller into the Event Log | This was implemented as a refactoring stage of the Replication system's code before a work on the new Ingest system began. | ||
Improve tracking of the asynchronous operations (jobs/requests) | |||
Garbage collection at Replication workers |
2. General code improvements, refactoring, cleanup
UNDER CONSIDERATION | PLANNED | STARTED | FINISHED |
These projects have generally low priority, and they're meant to improve the overall quality and robustness of the code, as well as make it more maintainable in the long run.
3. Known bugs, outstanding issues, etc.
- Investigating problems with XRootD/SSI at workers
- Huge memory leak in the Master Replication Controller when Qserv workers are down