You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
With this commit we leverage new functionality in Rally to spin up a
cluster before all challenges are executed. This allows us to run
challenges which depend on others (e.g. query only challenges that rely
on another challenge to actually put the data into the cluster).
Closeselastic#49
Copy file name to clipboardExpand all lines: smoke-test.sh
+43-7Lines changed: 43 additions & 7 deletions
Original file line number
Diff line number
Diff line change
@@ -24,21 +24,57 @@ set -u
24
24
# fail on pipeline errors, e.g. when grepping
25
25
set -o pipefail
26
26
27
+
readonly RACE_ID=$(uuidgen)
27
28
readonly ES_VERSION=${ES_VERSION:-7.3.0}
28
-
# intentionally not tested (at the moment) because these challenges require running a different challenge first against the same cluster:
29
+
30
+
# the order matters here as later challenges might expect that some indices already exist that have been created by challenges running earlier. In particular, these challenges are affected:
29
31
#
30
32
# * frozen-querying (depends on frozen-data-generation)
31
33
# * combined-indexing-and-querying (depends on any challenge that has already created elasticlogs-q* indices)
32
34
# * elasticlogs-querying (depends on any challenge that has already created elasticlogs-q* indices)
0 commit comments