4823: Switched order of diff (now gives a diff *to* the target collection

[arvados.git] / sdk / python / arvados / keep.py

diff --git a/sdk/python/arvados/keep.py b/sdk/python/arvados/keep.py
index a087838d261d59d0d43eaa252693cf041f3f4a94..d1b07cce0d558eff9a13a8ce57026718bfd6a4a3 100644 (file)
--- a/sdk/python/arvados/keep.py
+++ b/sdk/python/arvados/keep.py
@@ -326,7 +326,7 @@ class KeepClient(object):
                 content = result.content
                 _logger.info("%s response: %s bytes in %s msec (%.3f MiB/sec)",
                              self.last_status(), len(content), t.msecs,
-                             (len(content)/(1024.0*1024))/t.secs)
+                             (len(content)/(1024.0*1024))/t.secs if t.secs > 0 else 0)
                 if self.success_flag:
                     resp_md5 = hashlib.md5(content).hexdigest()
                     if resp_md5 == locator.md5sum:
@@ -663,17 +663,27 @@ class KeepClient(object):
         if loop.success():
             return blob
 
-        # No servers fulfilled the request.  Count how many responded
-        # "not found;" if the ratio is high enough (currently 75%), report
-        # Not Found; otherwise a generic error.
+        try:
+            all_roots = local_roots + hint_roots
+        except NameError:
+            # We never successfully fetched local_roots.
+            all_roots = hint_roots
         # Q: Including 403 is necessary for the Keep tests to continue
         # passing, but maybe they should expect KeepReadError instead?
-        not_founds = sum(1 for ks in roots_map.values()
-                         if ks.last_status() in set([403, 404, 410]))
-        if roots_map and ((float(not_founds) / len(roots_map)) >= .75):
-            raise arvados.errors.NotFoundError(loc_s)
+        not_founds = sum(1 for key in all_roots
+                         if roots_map[key].last_status() in {403, 404, 410})
+        service_errors = ((key, roots_map[key].last_result)
+                          for key in all_roots)
+        if not roots_map:
+            raise arvados.errors.KeepReadError(
+                "failed to read {}: no Keep services available ({})".format(
+                    loc_s, loop.last_result()))
+        elif not_founds == len(all_roots):
+            raise arvados.errors.NotFoundError(
+                "{} not found".format(loc_s), service_errors)
         else:
-            raise arvados.errors.KeepReadError(loc_s)
+            raise arvados.errors.KeepReadError(
+                "failed to read {}".format(loc_s), service_errors)
 
     @retry.retry_method
     def put(self, data, copies=2, num_retries=None):
@@ -734,9 +744,17 @@ class KeepClient(object):
 
         if loop.success():
             return thread_limiter.response()
-        raise arvados.errors.KeepWriteError(
-            "Write fail for %s: wanted %d but wrote %d" %
-            (data_hash, copies, thread_limiter.done()))
+        if not roots_map:
+            raise arvados.errors.KeepWriteError(
+                "failed to write {}: no Keep services available ({})".format(
+                    data_hash, loop.last_result()))
+        else:
+            service_errors = ((key, roots_map[key].last_result)
+                              for key in local_roots
+                              if not roots_map[key].success_flag)
+            raise arvados.errors.KeepWriteError(
+                "failed to write {} (wanted {} copies but wrote {})".format(
+                    data_hash, copies, thread_limiter.done()), service_errors)
 
     # Local storage methods need no-op num_retries arguments to keep
     # integration tests happy.  With better isolation they could