> On Jan. 21, 2015, 11:11 a.m., Timothy Chen wrote:
> > src/slave/containerizer/fetcher.cpp, line 452
> > <https://reviews.apache.org/r/30037/diff/5/?file=827816#file827816line452>
> >
> >     What is the reason you want to get the extension?
> >     I see you're using it at the cache file name, but I don't understand 
> > why is that necessary?
> >     
> >     Also If we pass a relative path (valid with MESOS_FRAMEWORK_HOME), then 
> > we no longer have a extension too.

Added a comment what the extension is for: 

  // Why we add the file extension here: When fetching from cache, if 
  // extraction is enabled, the extraction algorithm can look at the 
  // extension of the cache file the same way as it would at a 
  // download of the original URI, and external commands performing 
  // the extraction do not get confused by their source file
  // missing an expected form of extension.


> On Jan. 21, 2015, 11:11 a.m., Timothy Chen wrote:
> > src/slave/containerizer/fetcher.cpp, line 456
> > <https://reviews.apache.org/r/30037/diff/5/?file=827816#file827816line456>
> >
> >     Why do we need to have a distinct cache file name? And also why 
> > introduce a PREFIX for it too? Are we expecting to distinguish files that 
> > is cache and isn't?
> >     
> >     And if we are doing so, I don't see why not uris can also start with c 
> > if they are relative paths.

Adding comments:

// Use a distinct cache file name for every URI. Different URIs
// may have the same base name and we do need to segregate the
// download results. This can be done by separate directories
// or by different file names. Since there may be tighter limits
// regarding how many sub-directories a file system can bear than
// on the number of files in a directory, we opt for the latter.

  // We put a prefix before the serial number just in case the
  // underlying file system does not accept digits as first file
  // name character.


- Bernd


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/30037/#review68964
-----------------------------------------------------------


On Jan. 21, 2015, 12:54 a.m., Bernd Mathiske wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/30037/
> -----------------------------------------------------------
> 
> (Updated Jan. 21, 2015, 12:54 a.m.)
> 
> 
> Review request for mesos, Adam B, Benjamin Hindman, Till Toenshoff, and 
> Timothy Chen.
> 
> 
> Bugs: MESOS-2069
>     https://issues.apache.org/jira/browse/MESOS-2069
> 
> 
> Repository: mesos-git
> 
> 
> Description
> -------
> 
> Extends the fetcher info protobuf with "actions" (fetch directly, through 
> cache, retrieve from cache). Switches the basis for dealing with uris to 
> "items", which contain the uri, the action, and potentially and cache file 
> name. Refactors fetch() and run(), so there is only one of each. Introduces 
> about half of the actual cache logic, including a hashmap of cache file 
> objects for bookkeeping and basic operations on it. 
> 
> 
> Diffs
> -----
> 
>   include/mesos/fetcher/fetcher.proto 
> facb87b92bf3194516f636dcc348e136af537721 
>   src/launcher/fetcher.cpp fed0105946da579a38357a30e7ae56e646e05b89 
>   src/slave/containerizer/fetcher.hpp 
> 1db0eaf002c8d0eaf4e0391858e61e0912b35829 
>   src/slave/containerizer/fetcher.cpp 
> 5993670f7899233efa1e6acef4b0c7856e32f748 
> 
> Diff: https://reviews.apache.org/r/30037/diff/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Bernd Mathiske
> 
>

Reply via email to