Fast input/output in a content-addressable storage architecture with paged metadata
An aspect of minimizing read amplification IO where metadata is not in RAM includes reading a l_md_page and corresponding lp_md_page from a storage device in a dual distributed layered architecture. The l_md_page specifies a metadata page that persists in a SSD and having logical addresses of metadata, and the lp_md_page associates logical block addresses with corresponding physical locations for the metadata. An aspect further includes reading data for a redundant array of independent disks (RAID) stripe according to an associated physical offset in the lp_md_page, accessing a stripe counter from the lp_md_page, and comparing the stripe counter from the lp_md_page to a stripe counter held in memory. Upon determining the stripe counter from the lp_md_page is the same, an aspect further includes determining the data is valid and reading the data according to associated physical offset in the lp_md_page and while bypassing a data module for the data.
1. A method for minimizing read amplification for input/output (IO) where metadata is not in random access memory (RAM), comprising:
reading a l_md_page and corresponding lp_md_page from a storage device in a dual distributed layered architecture, the l_md_page specifies a metadata page that persists in a solid state drive and having logical addresses of metadata, and the lp_md_page associates logical block addresses (LBAs) with corresponding physical locations for the metadata;
concatenating the l_md_page and the lp_md_page;
storing the concatenated pages in the solid state disk;
wherein reading the reading the l_md_page and the lp_md_page includes loading the concatenated pages in a single command;
reading data for a redundant array of independent disks (RAID) stripe according to an associated physical offset in the lp_md_page;
accessing a stripe counter from the lp_md_page;
comparing the stripe counter from the lp_md_page to a stripe counter held in memory; and
upon determining the stripe counter from the lp_md_page is the same:
determining the data is valid; and
reading the data according to associated physical offset in the lp_md_page and while bypassing a data module for the data.
2. The method of claim 1 , wherein reading the l_md_page and the lp_md_page is performed by sending an input/output (IO) request to the data module.
3. The method of claim 1 , wherein reading the l_md_page and the lp_md_page is performed by a direct non-volatile memory over fabric (NVMEoF) read from a control module to drives bypassing the data module, the reading performed if the data is determined to be correct.
4. The method of claim 1 , wherein accessing the stripe counter is performed by sending a request to a data module.
5. The method of claim 1 , wherein upon accessing the stripe counter the method includes reading a remote direct memory access (RDMA) read via an RDMA interconnect from relevant offset in mapped memory.
6. The method of claim 1 , further comprising upon determining the
stripe counter from the lp_md_page is not the same:
accessing a data module and loading the h_md_page;
reading a physical location of the data from the h_md_page;
accessing by a control module, a new physical location of the data during a read operation; and
updating, by the control module, the lp_md_page.
7. A system for minimizing read amplification for input/output (IO) where metadata is not in random access memory (RAM), comprising:
a memory comprising computer-executable instructions; and
a processor executing the computer-executable instructions, the computer-executable instructions when executed by the processor cause the processor to perform operations comprising:
reading a l_md_page and corresponding lp_md_page from a storage device in a dual distributed layered architecture, the l_md_page specifies a metadata page that persists in a solid state drive and having logical addresses of metadata, and the lp_md_page associates logical block addresses (LBAs) with corresponding physical locations for the metadata;
concatenating the l_md_page and the lp_md_page;
storing the concatenated pages in the solid state disk;
wherein reading the reading the l_md_page and the lp_md_page includes loading the concatenated pages in a single command;
reading data for a redundant array of independent disks (RAID) stripe according to an associated physical offset in the lp_md_page;
accessing a stripe counter from the lp_md_page;
comparing the stripe counter from the lp_md_page to a stripe counter held in memory; and
upon determining the stripe counter from the lp_md_page is the same:
determining the data is valid; and
reading the data according to associated physical offset in the lp_md_page and while bypassing a data module for the data.
8. The system of claim 7 , wherein reading the l_md_page and the lp_md_page is performed by sending an input/output (IO) request to the data module.
9. The system of claim 7 , wherein reading the l_md_page and the lp_md_page is performed by a direct NVMEoF read from a control module to drives bypassing the data module, the reading performed if the data is determined to be correct.
10. The system of claim 7 , wherein accessing the stripe counter is performed by sending a request to a data module.
11. The system of claim 7 , wherein upon accessing the stripe counter, the operations further include reading an RDMA read via an RDMA interconnect from relevant offset in mapped memory.
12. The system of claim 7 , wherein the operations further include upon determining the stripe counter from the lp_md_page is not the same:
accessing a data module and loading the h_md_page;
reading a physical location of the data from the h_md_page;
accessing by a control module, a new physical location of the data during a read operation; and
updating, by the control module, the lp_md_page.
13. A computer program product embodied on a non-transitory computer readable medium, the computer program product including instructions that, when executed by a computer causes the computer to perform operations comprising:
reading a l_md_page and corresponding lp_md_page from a storage device in a dual distributed layered architecture, the l_md_page specifies a metadata page that persists in a solid state drive and having logical addresses of metadata, and the lp_md_page associates logical block addresses (LBAs) with corresponding physical locations for the metadata;
concatenating the l_md_page and the lp_md_page;
storing the concatenated pages in the solid state disk;
wherein reading the reading the l_md_page and the lp_md_page includes loading the concatenated pages in a single command:
reading data for a redundant array of independent disks (RAID) stripe according to an associated physical offset in the lp_md_page;
accessing a stripe counter from the lp_md_page;
comparing the stripe counter to from the lp_md_page to a stripe counter held in memory; and
upon determining the stripe counter from the lp_md_page is the same:
determining the data is valid; and
reading the data according to associated physical offset in the lp_md_page and while bypassing a data module for the data.
14. The computer program product of claim 13 , wherein reading the l_md_page and the lp_md_page is performed by sending an input/output (IO) request to the data module.
15. The computer program product of claim 13 , wherein reading the l_md_page and the lp_md_page is performed by a direct NVMEoF read from a control module to drives bypassing the data module, the reading performed if the data is determined to be correct.
16. The computer program product of claim 13 , wherein accessing the stripe counter is performed by sending a request to a data module.
17. The computer program product of claim 13 , wherein upon accessing the stripe counter the operations further include reading an RDMA read via an RDMA interconnect from relevant offset in mapped memory.
18. The computer program product of claim 13 , wherein the operations further include:
upon determining the stripe counter from the lp_md_page is not the same:
accessing a data module and loading the h_md_page;
reading a physical location of the data from the h_md_page;
accessing by a control module, a new physical location of the data during a read operation; and
updating, by the control module, the lp_md_page.