Collaboration with LTSI Testing

LTSI Workshop 2014 at ELC

Transcript

1. Copyright 2014, Toshiba Corporation. Collaboration with LTSI Testing Yoshitake Kobayashi Corporate Software Engineering Center Toshiba Corporation
2. 2 LTSI Workshop / Embedded Linux Conference 2014    
3. 3 LTSI Workshop / Embedded Linux Conference 2014   
4. 4 LTSI Workshop / Embedded Linux Conference 2014        
5. 5 LTSI Workshop / Embedded Linux Conference 2014                  
6. 6 LTSI Workshop / Embedded Linux Conference 2014     Patches available on upstream Patches available on Xilinx’s repository LTSI-3.10 backport backport
7. 7 LTSI Workshop / Embedded Linux Conference 2014                 
8. 8 LTSI Workshop / Embedded Linux Conference 2014                  
9. 9 LTSI Workshop / Embedded Linux Conference 2014     
10. 10 LTSI Workshop / Embedded Linux Conference 2014 Target files Writer processes (N procs) Target Host write() system call Log Host Logger Each writer process • writes to text files (ex. 100 files) • sends progress log to logger This test case available at the following URL: https://github.com/ystk/fs-test Reset at random timing
11. 11 LTSI Workshop / Embedded Linux Conference 2014 Checker Target file LOG file AAAAA BBBBB CCCCC DDDDD EEEEE OK FFFFF AAAAA BBBBB CCCCC DDDDD EEEEE OK AAAAA BBBBB CCCCC DDDDD AAAAA NG AAAAA BBBBB CCCCC DDDDD NG ? size mismatch data mismatch Verify the following metrics • file size • file contents Estimated file contents Estimated file size
12. 12 LTSI Workshop / Embedded Linux Conference 2014  0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL * * * JFS XFS Error rate [%] 45.9% Error rate [%] 0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL EXT4- JOURNAL EXT4- ORDERED EXT4- WRITEBACK * XFS BTRFS 82.4% 84.7% 43.4% 41.4% 43.2% Error rate [%] 0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL EXT4- JOURNAL EXT4- ORDERED * JFS XFS 13.3% kernel 2.6.18 kernel 2.6.31 kernel 2.6.33 File size mismatch Data mismatch
13. 13 LTSI Workshop / Embedded Linux Conference 2014  0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL * * * JFS XFS Error rate [%] 45.9% Error rate [%] 0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL EXT4- JOURNAL EXT4- ORDERED EXT4- WRITEBACK * XFS BTRFS 82.4% 84.7% 43.4% 41.4% 43.2% Error rate [%] 0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL EXT4- JOURNAL EXT4- ORDERED * JFS XFS 13.3% kernel 2.6.18 kernel 2.6.31 kernel 2.6.33 File size mismatch Data mismatch Point 1： A file system has different characteristics of data reliability
14. 14 LTSI Workshop / Embedded Linux Conference 2014  0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL * * * JFS XFS Error rate [%] 45.9% Error rate [%] 0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL EXT4- JOURNAL EXT4- ORDERED EXT4- WRITEBACK * XFS BTRFS 82.4% 84.7% 43.4% 41.4% 43.2% Error rate [%] 0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL EXT4- JOURNAL EXT4- ORDERED * JFS XFS 13.3% kernel 2.6.18 kernel 2.6.31 kernel 2.6.33 File size mismatch Data mismatch Point 1： A file system has different characteristics of data reliability Point 2: Some Results depends on kernel version
15. 15 LTSI Workshop / Embedded Linux Conference 2014  0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL * * * JFS XFS Error rate [%] 45.9% Error rate [%] 0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL EXT4- JOURNAL EXT4- ORDERED EXT4- WRITEBACK * XFS BTRFS 82.4% 84.7% 43.4% 41.4% 43.2% Error rate [%] 0.00 0.50 1.00 1.50 2.00 EXT3- ORDERED EXT3- JOURNAL EXT4- JOURNAL EXT4- ORDERED * JFS XFS 13.3% kernel 2.6.18 kernel 2.6.31 kernel 2.6.33 File size mismatch Data mismatch Point 1： A file system has different characteristics of data reliability Point 2: Some Results depends on kernel version Point 3: EXT4-Journal and BTRFS has a nice result
16. 16 LTSI Workshop / Embedded Linux Conference 2014  START GOAL Accelerate Things that cannot be accelerated • CPU clock • I/O access speed (ex. SSD) • Network bandwidth • etc. Try to detect errors that caused by clock Focus to accelerate clock Issues • Long-term testing takes really long time → We want results as fast as possible
17. 17 LTSI Workshop / Embedded Linux Conference 2014  Xdaliclock works as a stopwatch Returned an incorrect value after about 450 days. (It takes about 6 hours in 1000 times acceleration)
18. 18 LTSI Workshop / Embedded Linux Conference 2014      
19. 19 LTSI Workshop / Embedded Linux Conference 2014                
20. 20 LTSI Workshop / Embedded Linux Conference 2014       
21. 21 LTSI Workshop / Embedded Linux Conference 2014              
22. 22 LTSI Workshop / Embedded Linux Conference 2014