diff --git a/coverage.sh b/coverage.sh deleted file mode 100755 index aff1a7b..0000000 --- a/coverage.sh +++ /dev/null @@ -1,9 +0,0 @@ -#!/bin/bash - -# Runs the coverage -mkdir -p build/Coverage -cd build/Coverage -cmake ../.. -DCMAKE_BUILD_TYPE=Coverage -make -./tester -make coverage diff --git a/example/benchmark/README.md b/example/benchmark/README.md index 0f0c72d..ea97a7f 100644 --- a/example/benchmark/README.md +++ b/example/benchmark/README.md @@ -1,9 +1,10 @@ # AnyBlobBenchmark -This directory contains the AnyBlob benchmarking tools, and also includes all third-parties (e.g., AWS SDK) to run our experiments for Section 2 and Section 3 of the AnyBlob paper. -The sub-directory scripts contains the scripts that we used to collect our raw data for the plots. +This directory contains the AnyBlob benchmarking tools, and also includes all third-parties (e.g., AWS SDK) to run our experiments of the AnyBlob paper. # Validate Experiments +## AnyBlob Experiments + To validate our findings you need to set-up an AWS storage account to create randomly generated data objects. Further, set up an instance with IAM Storage permissions and copy the files to this node. For throughput experiments, we recommend the c5n.18xlarge, but a cheaper instance such as the c5n.xlarge can be used for latency experiments. @@ -16,10 +17,10 @@ The raw data of our Figures were created from multiple scripts. This list contai - sizes_latency.sh for Figure 2 (Section 2.3) - latency.sh for Figure 3 (Section 2.3) - latency_sparse.sh for Figure 4 (Section 2.3) -- throughput_sparse.sh for Figures 5 & 6 (Section 2.4) -- sizes.sh for Figure 7 (Section 2.5) -- model.sh for Section 2.8 and Figure 9 (Section 3.4) -- https.sh for Section 2.7 +- throughput_sparse.sh for Figures 5, 6 & 7 (Section 2.4) +- sizes.sh for Figure 8 (Section 2.5) +- https.sh for Figure 9 (Section 2.6) +- model.sh for Figure 10 (Section 2.8) and Figure 12 (Section 3.4) In the following, we show the steps to compile our benchmarking binary. @@ -31,3 +32,6 @@ In the following, we show the steps to compile our benchmarking binary. cmake ../.. -DCMAKE_BUILD_TYPE=Release make -j16 +## Database Experiments + +All our DBMS related experiments and our binary of our proprietary system are shared in https://gitlab.db.in.tum.de/durner/cloud-storage-analytics.