Live data from Hacker News

Viewing profile — sghemawat

sghemawat

HN member
Joined
Sun, May 08, 2011, 5:14 PM UTC
HN karma
268
Public activity
9 items

About sghemawat

Sanjay Ghemawat

Recent public activity

  1. comment
    Comment #46549761

    Hi Kenton! No worries at all. I tend to be quieter than Jeff anyway (less public speaking etc.) and I am happy to not have a dedicated website. :-). -Sanjay

  2. comment
    Comment #5263221

    Not your fault. We just plain forgot to update the documentation, so the freshest available document is a few years out of date.

  3. comment
    Comment #5263159

    About [1] Sorry about that: the document you linked to is amazingly stale. tcmalloc has been releasing memory to the system for many years. See for example the IncrementalScavenge …

  4. comment
    Comment #2527623

    We have no such plans. The library is designed so that concurrency control is outside its scope. If I needed it for an application, I might consider using some application level co…

  5. comment
    Comment #2527611

    The use of c++0x should be limited to the small AtomicPointer class in port_posix.h. The class uses for its implementation. If you can provide a different implementation of "acquir…

  6. comment
    Comment #2526799

    Merely an incomplete one. Leveldb never writes in place: it always appends to a log file, or merges existing files together to produce new ones. So an OS crash will cause a partial…

  7. comment
    Comment #2526394

    leveldb is a persistent ordered map; bitcask is a persistent hash table (no ordered iteration). bitcask stores a fixed size record in memory for every key. So for databases with la…

  8. comment
    Comment #2526355

    One of the leveldb authors here. TokyoCabinet is something we seriously considered using instead of writing leveldb. TokyoCabinet has great performance usually. I haven't done a ca…

  9. comment
    Comment #2526311

    It is durable from an OS perspective. I started typing in a long explanation, but I think the comment for the "sync" field in the options.h file captures things well: struct WriteO…