make jepsen test more robust: handle errors and timeouts, fixed access key

3 files changed, 36 insertions, 26 deletions

diff --git a/script/jepsen.garage/src/jepsen/garage.clj b/script/jepsen.garage/src/jepsen/garage.clj
index 5816512b..891fdf12 100644
--- a/script/jepsen.garage/src/jepsen/garage.clj
+++ b/script/jepsen.garage/src/jepsen/garage.clj
@@ -41,7 +41,7 @@
   [opts]
   (let [workload ((get workloads (:workload opts)) opts)
         garage-version (if (:increasing-timestamps opts)
-                         "03490d41d58576d7b3bcf977b2726d72a3a66ada"
+                         "d146cdd5b66ca1d3ed65ce93ca42c6db22defc09"
                          "v0.9.0")]
     (merge tests/noop-test
            opts
@@ -56,16 +56,23 @@
                                   (gen/stagger (/ (:rate opts)))
                                   (gen/nemesis
                                     (cycle [(gen/sleep 5)
-                                            {:type :info, :f :start}
+                                            ;{:type :info, :f :partition-start}
+                                            ;(gen/sleep 5)
+                                            {:type :info, :f :clock-scramble}
                                             (gen/sleep 5)
-                                            {:type :info, :f :stop}]))
+                                            ;{:type :info, :f :partition-stop}
+                                            ;(gen/sleep 5)
+                                            {:type :info, :f :clock-scramble}]))
                                   (gen/time-limit (:time-limit opts)))
                                 (gen/log "Healing cluster")
-                                (gen/nemesis (gen/once {:type :info, :f :stop}))
+                                (gen/nemesis (gen/once {:type :info, :f :partition-stop}))
                                 (gen/log "Waiting for recovery")
                                 (gen/sleep 10)
                                 (gen/clients (:final-generator workload)))
-            :nemesis          (nemesis/partition-random-halves)
+            :nemesis          (nemesis/compose
+                                {{:partition-start :start
+                                  :partition-stop :stop} (nemesis/partition-random-halves)
+                                 {:clock-scramble :scramble} (nemesis/clock-scrambler 20.0)})
             :checker          (checker/compose
                                 {:perf (checker/perf)
                                  :workload (:checker workload)})
diff --git a/script/jepsen.garage/src/jepsen/garage/daemon.clj b/script/jepsen.garage/src/jepsen/garage/daemon.clj
index 81163521..2dcfa94f 100644
--- a/script/jepsen.garage/src/jepsen/garage/daemon.clj
+++ b/script/jepsen.garage/src/jepsen/garage/daemon.clj
@@ -15,7 +15,8 @@
 (def pidfile (str base-dir "/garage.pid"))
 
 (def admin-token "icanhazadmin")
-(def access-key "jepsen")
+(def access-key-id "GK8bfb6a51286071c6c9cd8bc3")
+(def secret-access-key "b0be95f71c1c6f16858a9edf395078b75c12ecb6b1c03385c4ae92076e4994a3")
 (def bucket-name "jepsen")
 
 ; THE GARAGE DB
@@ -78,10 +79,10 @@
   (c/trace
     (c/exec binary :layout :apply :--version 1)
     (info node "garage status:" (c/exec binary :status))
-    (c/exec binary :key :create access-key)
+    (c/exec binary :key :import access-key-id secret-access-key :--yes)
     (c/exec binary :bucket :create bucket-name)
-    (c/exec binary :bucket :allow :--read :--write bucket-name :--key access-key)
-    (info node "key info: " (c/exec binary :key :info access-key))))
+    (c/exec binary :bucket :allow :--read :--write bucket-name :--key access-key-id)
+    (info node "key info: " (c/exec binary :key :info access-key-id))))
 
 (defn db
   "Garage DB for a particular version"
@@ -122,13 +123,9 @@
 (defn creds
   "Obtain Garage credentials for node"
   [node]
-  (let [key-info (c/on node (c/exec binary :key :info access-key :--show-secret))
-        [_ ak sk] (re-matches
-                    #"(?s).*Key ID: (.*)\nSecret key: (.*)\nCan create.*"
-                    key-info)]
-    {:access-key ak
-     :secret-key sk
-     :endpoint (str "http://" node ":3900")
-     :bucket bucket-name
-     :client-config {:path-style-access-enabled true}}))
+  {:access-key access-key-id
+   :secret-key secret-access-key
+   :endpoint (str "http://" node ":3900")
+   :bucket bucket-name
+   :client-config {:path-style-access-enabled true}})
 
diff --git a/script/jepsen.garage/src/jepsen/garage/reg.clj b/script/jepsen.garage/src/jepsen/garage/reg.clj
index 3a2c1d55..d5404c6b 100644
--- a/script/jepsen.garage/src/jepsen/garage/reg.clj
+++ b/script/jepsen.garage/src/jepsen/garage/reg.clj
@@ -33,19 +33,25 @@
     (let [[k v] (:value op)]
       (case (:f op)
         :read
-          (let [value (s3/get (:creds this) k)]
-            (assoc op :type :ok, :value (independent/tuple k value)))
+          (try+
+            (let [value (s3/get (:creds this) k)]
+              (assoc op :type :ok, :value (independent/tuple k value)))
+            (catch (re-find #"Unavailable" (.getMessage %)) ex
+              (assoc op :type :fail, :error [:s3-error (.getMessage ex)])))
         :write
-          (do
-            (s3/put (:creds this) k v)
-            (assoc op :type :ok)))))
+          (try+
+            (do
+              (s3/put (:creds this) k v)
+              (assoc op :type :ok))
+            (catch (re-find #"Unavailable" (.getMessage %)) ex
+              (assoc op :type :fail, :error [:s3-error (.getMessage ex)]))))))
   (teardown! [this test])
   (close! [this test]))
 
 (defn workload
   "Tests linearizable reads and writes"
   [opts]
-  {:client           (RegClient. nil)
+  {:client           (client/timeout 10 (RegClient. nil))
    :checker          (independent/checker
                        (checker/compose
                          {:linear (checker/linearizable
@@ -53,8 +59,8 @@
                                      :algorithm :linear})
                           :timeline (timeline/html)}))
    :generator        (independent/concurrent-generator
-                       10
-                       (range)
+                       (/ (:concurrency opts) 10) ; divide threads in 10 groups
+                       (range)                    ; working on 10 keys
                        (fn [k]
                          (->>
                            (gen/mix [op-get op-put op-del])