Just curious why are there two path style?

[YifangSun]

Here in common/source/common/toolkit/StorageTk.h

beegfs/common/source/common/toolkit/StorageTk.h

Lines 342 to 358 in d7fc807

	* @return
	* chunkPathV2 (no orig feature):
	* hashDir1/hashDir2/entryID
	* chunkPathV3:
	* uidX/level1/level2/parentID/entryID
	*/
	static std::string getFileChunkPath(const PathInfo* pathInfo, const std::string entryID)
	{
	std::string chunkPath;
	if (!pathInfo->hasOrigFeature() )
	{ // 2012.10 style hash dir
	chunkPath = getFileChunkPathV2(entryID, NULL); // old storage format
	}
	else
	{ // 2014.01 style path with user and timestamp dirs

Why you guys switch to a new version of chunk path?

Are there any technical reason? Like concurrency preformance? Or create file latency? Or management consideration?

@philippfalk @iamjoemccormick Thanks!

[iamjoemccormick]

Hi @YifangSun ,

Thanks a lot for your interest in BeeGFS! The codebase is public and we encourage exploration, but for maintainers to spend time on Q&A we need to understand the background (bug report, contribution idea, etc.). If you can provide that, we’ll do our best to help point you in the right direction.

That said, the general answer here is that this change was introduced in version 2014.01. From the original release notes:

• New Data Layout: Storage servers now use a new data layout with user- and
  time-based grouping of chunk files. The new layout significantly reduces disk
  seeks by improving cache efficiency of the underlying local server file system
  and helps to avoid aging-related performance reduction.

[YifangSun]

Thanks for your reply!

[YifangSun]

Why don't meta apply the same change? Looks like meta is still using 2 layers hash dir.