Comment 2 for bug 512216

Revision history for this message
Eli Zaretskii (eliz) wrote :

Attached below. My net connection dropped during that command (would you believe that?), and bzr errored out, but the index update was working for about 20 minutes up until that moment, so I hope the callgrind file has enough information for you. FWIW, most of the time is spent in this phase:

  Indexing...:Indexing commits 2/4

One of the cores is almost 100% busy during all this time, and there's also significant (but not overwhelming) disk I/O.