[bug] move assign to prevent race condition

When the error happens too quickly, the errback is called before the
assign of the callLater. And in the errback we cancel that call which
gives an error.

- Related: #7451

1 files changed, 2 insertions, 2 deletions

diff --git a/src/leap/bitmask/services/soledad/soledadbootstrapper.py b/src/leap/bitmask/services/soledad/soledadbootstrapper.py
index ec758833..f3d50925 100644
--- a/src/leap/bitmask/services/soledad/soledadbootstrapper.py
+++ b/src/leap/bitmask/services/soledad/soledadbootstrapper.py
@@ -623,10 +623,10 @@ class Syncer(object):
         logger.debug("BOOTSTRAPPER: trying to sync Soledad....")
         # pass defer_decryption=False to get inline decryption
         # for debugging.
-        self._sync_deferred = self._soledad.sync(defer_decryption=True)
-        self._sync_deferred.addCallbacks(self._success, self._error)
         self._timeout_delayed_call = reactor.callLater(self.WAIT_MAX_SECONDS,
                                                        self._timeout)
+        self._sync_deferred = self._soledad.sync(defer_decryption=True)
+        self._sync_deferred.addCallbacks(self._success, self._error)
 
     def _success(self, result):
         logger.debug("Soledad has been synced!")