There is a long discussion here[1] but short answers are
1) The cache key is hashed and that value used to select a stripe (span).
2) It's a circular buffer so the first 50M are written at the end and the next
150M at the beginning after wrapping. See[2] for more detail, scroll down to
where a paragraph starts "Large objects..."
[1] Cache Architecture — Apache Traffic Server 6.2.0 documentation[2] Cache
Architecture — Apache Traffic Server 6.2.0 documentation
|
| |
Cache Architecture — Apache Traffic Server 6.2.0 documentation
| |
|
|
| |
Cache Architecture — Apache Traffic Server 6.2.0 documentation
| |
|
On Wednesday, April 13, 2016 7:42 AM, Eric Friedrich (efriedri)
<[email protected]> wrote:
Hello,
Two questions on ATS cache write logic
1) How does ATS balance write load across different storage devices?
(Round-robin, least I/O load, etc??)
2) How does ATS handle caching large files near the end of cache spans? For
example, I want to cache a 200MB object but my write pointer is only 50MB from
the end of the cyclone buffer? I know ATS will not split files across cache
spans, so will it skip the last 50MB and wrap back to the beginning of the
buffer?
Thanks,
Eric
