ehcache

Checkout Tools
  • last updated a few seconds ago
Constraints
Constraints: committers
 
Constraints: files
Constraints: dates
core is relocated now

    • -86
    • +0
    ./distribution/MulticastRMIPeerProviderPerfTest.java
    • -104
    • +0
    ./CacheClassicLruMemoryStorePerfTest.java
    • -65
    • +0
    ./constructs/blocking/CountingCachePerfEntryFactory.java
    • -173
    • +0
    ./DiskStorePerformancePerfTest.java
    • -118
    • +0
    ./distribution/PayloadUtilPerfTest.java
    • -35
    • +0
    ./store/LruMemoryStorePerfTest.java
  1. … 1097 more files in changeset.
MNK-3026 : Element serialized size changed
    • -1
    • +1
    ./distribution/RMICacheReplicatorPerfTest.java
Rolling back DEV-6081 changes as requested by Steve
Fix 6081
  1. … 1 more file in changeset.
fix test

    • -0
    • +46
    ./store/disk/PerfDiskStoreHelper.java
fix tests

    • -9
    • +14
    ./distribution/MulticastRMIPeerProviderPerfTest.java
    • -17
    • +24
    ./distribution/RMICacheReplicatorPerfTest.java
fix tests

    • -10
    • +2
    ./constructs/blocking/SelfPopulatingCachePerfTest.java
fix test

    • -4
    • +2
    ./ConcurrencyProblemCachePerfTest.java
updated sizes to reflect what the sizeof engine measures
Record perf test runs
  1. … 1 more file in changeset.
MNK-2613 I don't see why this test uses an overflowing cache. Mem footprint would be bigger. Making that a non overflowing one
Added some information when failing
updated assertion to reflect new Element serialized size
    • -1
    • +1
    ./distribution/RMICacheReplicatorPerfTest.java
updated tests to match new tiered disk store behavior
    • -0
    • +46
    ./CacheClassicLruMemoryStorePerfTest.java
  1. … 1 more file in changeset.
fixed tests
    • -11
    • +8
    ./store/LfuMemoryStorePerfTest.java
Some changes to perf tests

- Ignore the tiny timeout test, it provides no extra coverage and fails sporadically for a non failure reason

- Switch to using System.nanoTime() for timing so we are not affected by system clock shifts (due to ntp?)

- Stop asserting on liveness - it has no meaning now that we don't use synchronization on the Cache instance at all

MNK-2037 : make sure testLowest increments hit counts correctly, add some explanatory comments
    • -44
    • +46
    ./store/LfuMemoryStorePerfTest.java
DEV-4430: Make ehcache perfTests not use /tmp for diskStore
  1. … 8 more files in changeset.
Beanshellio
    • -1
    • +1
    ./distribution/RMICacheReplicatorPerfTest.java
  1. … 3 more files in changeset.
DEV-4187 DEV-4188 Moved all perf tests to dedicated perfTests profile, marked all flaky tests with TODO: FIXME and added some notes
    • -0
    • +81
    ./distribution/MulticastRMIPeerProviderPerfTest.java
    • -0
    • +61
    ./CacheManagerPerfTest.java
    • -0
    • +58
    ./CacheClassicLruMemoryStorePerfTest.java
    • -0
    • +65
    ./constructs/blocking/CountingCachePerfEntryFactory.java
    • -0
    • +173
    ./DiskStorePerformancePerfTest.java
    • -0
    • +118
    ./distribution/PayloadUtilPerfTest.java
    • -0
    • +237
    ./DiskStorePerfTest.java
    • -0
    • +691
    ./CachePerfTest.java
    • -0
    • +35
    ./store/LruMemoryStorePerfTest.java
    • -0
    • +513
    ./distribution/RMICacheReplicatorPerfTest.java
  1. … 49 more files in changeset.