- Trace format:
- column 1: It's Chinese words, which means AM or PM. You can ignore it.
- column 2: Access time in the format: hh:mm:ss
- column 3: R means read opeartion. W means write operation.
- column 4: The starting LBA being accessed
- column 5: The number of continuous LBAs being accessed in this R/W opeartion.
- Major Contributor: Dr. Li-Pin Chang @ the NEWS Lab of NTU
-
The traces of the trace-driven access pattern were collected
by emulating web-surfing applications over a portable device,
and the characteristics of the traces are shown in the following
table:
| Attribute |
Value |
| File system |
FAT32 |
| Running Applications |
Web surfing, emails sending
/ receiving, movie playing and
downloading, document
typesetting, and gaming. |
| Sector Size/Page Size |
512 Bytes |
| Duration |
One month |
| Final Capacity Utilization |
71% (Initially Empty) |
| Total Data Written |
18.284 MB |
| Read / Write Ratio |
48% / 52% |
| Mean Read Size |
8.2 Sectors |
| Mean Write Size |
5.7 Sectors |
| Inter-Arrival Time |
Mean: 32 ms
Standard Deviation: 229 ms |
| Locality |
74-26 (74% of Total Requests Access
26% of Total LBA's) |
| Length of the Two-Level LRU List |
512 (Hot) / 1024 (Candidate) |
- If you use this trace in your paper, please cite the following paper:
- Li-Pin Chang and Tei-Wei Kuo, "An Adaptive Stripping Architecture for Flash Memory Storage Systems of Embedded Systems," IEEE Eighth Real-Time and Embedded Technology and Applications Symposium (RTAS), San Jose, USA, Sept 2002.
|
- Two geographic files (BUILDING.SHP and ROAD.SHP) describing
the roads and buildings of Taipei map were adopted as the data
sets for the experiments. Note that BUILDING.SHX and ROAD.SHX
are two index files to quickly retrieve the location information
of any object stored in BUILDING.SHP and ROAD.SHP. Note that
the geographic files were in the shapefile format. The numbers
of spatial objects of the buildings and roads are 8,588 and
7,338, respectively.
- Each line in these trace files conforms to the following
format:
- [I or M] ID xmin ymin xmax ymax
e.g., I 3 10 20 50 100 The above example is to insert an object
whose ID is 3 and the bounding box's coordinates are (10, 20)
and (50, 100).
- Note that I and M denote a insertion and a modification to
a object whose bounding box's coordinates are (xmin,ymin) and
(xmax,ymax). xmin, ymin, xmax, and ymax are in interval [0,100000].
We use these bounding boxes to create and maintain an R-Tree
index structure. We describe the trace files in the following:
| File Name |
Description |
building_sequential.txt
road_sequential.txt
|
The objects are sequentially inserted to an
R-Tree index structure according to their object IDs. |
building_random.txt
road_random.txt
|
The objects are randomly inserted to an R-Tree
index structure. |
building_modify_2.txt
road_modify_2.txt
|
The objects are sequentially inserted to an
R-Tree index structure according to their object IDs. Then,
20% of the inserted objects are randomly modified. |
building_modify_4.txt
road_modify_4.txt
|
The objects are sequentially inserted to an
R-Tree index structure according to their object IDs. Then,
40% of the inserted objects are randomly modified. |
building_modify_6.txt
road_modify_6.txt
|
The objects are sequentially inserted to an
R-Tree index structure according to their object IDs. Then,
60% of the inserted objects are randomly modified. |
building_modify_8.txt
road_modify_8.txt
|
The objects are sequentially inserted to an
R-Tree index structure according to their object IDs. Then,
80% of the inserted objects are randomly modified. |
- Reference:
1. ESRI Shapefile Technical Description
www.esri.com/library/whitepapers/pdfs/shapefile.pdf
2. C. H. Wu, L. P. Chang, and T. W. Kuo, ``An Efficient R-Tree
Implementation over Flash-Memory Storage Systems,'' The 11th
International Symposium on Advances in Geographic Information
Systems (ACM-GIS 2003).
|
| |
|
- Trace format:
- column 1: Access number
- column 2: It's Chinese words, which means AM or PM. You can ignore it.
- column 3: Access time in the format: hh:mm:ss
- column 4: Access duration
- column 5: Disk number
- column 6: "Read" means read opeartion. "Write" means write operation.
- column 7: The starting LBA being accessed
- column 8: The number of continuous LBAs being accessed in this R/W opeartion.
- Major Contributor: Joshon @ the NEWS Lab of NTU
-
The traces of the trace-driven access pattern were collected
by general PC applications over a personal computer,
and the characteristics of the traces are shown in the following
table:
| Attribute |
Value |
| File system |
NTFS |
| Running Applications |
Web surfing, emails sending
/ receiving, movie playing and
downloading, document
typesetting, and gaming. |
| Sector Size/Page Size |
512 Bytes |
| Duration |
One month |
|
