├── Delta-net_Interface.pdf └── README.md /Delta-net_Interface.pdf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/delta-net/datasets/9c90d989c9a2a68b6c642c829291bc02e5aacb61/Delta-net_Interface.pdf -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Delta-net 2 | 3 | Delta-net is a tool that makes it possible to detect and prevent certain network 4 | outages before they occur. For this, Delta-net analyzes changes to the data plane 5 | in real-time and incrementally maintains flow information of every packet in the 6 | entire network. In our NSDI'17 paper, we propose the first provably quasi-linear 7 | algorithm to do so. Here, we provide the data sets that we used in our experiments. 8 | 9 | If you are interested to learn more, please check out [our publications](#publications). 10 | For any questions, feel also free to contact us. 11 | 12 | ## Data Sets 13 | 14 | Our [NSDI'17 data set][data-set-nsdi-17] comprise several hundred million IP prefix 15 | rules generated using topologies and BGP updates from real-world deployed networks, 16 | including a globally deployed [ONOS][onos] software-defined networking application, 17 | called [SDN-IP][sdn-ip]. These data sets are organized as text files in which line 18 | denotes an insertion or removal of a rule. The format of such lines is explained 19 | in a short [manual][manual] that describes Delta-net's interface. 20 | 21 | Our experiments show that Delta-net checks a rule insertion and removal in tens of 22 | microseconds on average, a more than 10X improvement over the state-of-the-art. We 23 | also show that Delta-net eliminates an inherent bottleneck in the state-of-the-art 24 | that restricts its use in answering Datalog-style "what if" queries. 25 | 26 | In our ICNP paper (see info below), we generalized Delta-net's lattice-theoretical 27 | analysis to handle multi-dimensional match conditions, e.g., ip-table rules. 28 | 29 | [sdn-ip]: https://wiki.onosproject.org/display/ONOS/SDN-IP+Architecture 30 | [onos]: http://onosproject.org/ 31 | [data-set-nsdi-17]: https://www.kaggle.com/alikherad/deltanet-nsdi17-dataset 32 | [manual]: ./Delta-net_Interface.pdf 33 | 34 | ## Publications 35 | 36 | ``` 37 | @inproceedings{HKP2017, 38 | author = {Horn, Alex and Kheradmand, Ali and Prasad, Mukul R.}, 39 | title = {Delta-net: Real-time Network Verification Using Atoms}, 40 | booktitle = {NSDI}, 41 | year = {2017}, 42 | } 43 | @inproceedings{HKP2019, 44 | author = {Horn, Alex and Kheradmand, Ali and Prasad, Mukul R.}, 45 | title = {A Precise and Expressive Lattice-theoretical Framework for Efficient Network Verification}, 46 | booktitle = {ICNP}, 47 | year = {2019}, 48 | } 49 | ``` 50 | 51 | ## Notice 52 | 53 | ``` 54 | THIS DATASET IS PROVIDED "AS IS" BY THE PREPARER OF THE DATA SET. BY USING THE 55 | DATASET, YOU EXPRESSLY UNDERSTAND AND AGREE THAT, EXCEPT TO THE EXTENT 56 | PROHIBITED BY LAW, ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 57 | TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 58 | PURPOSE OR SPECIFIC PERFORMANCE USING SUCH DATASET AND/OR ANY WARRANTIES THAT 59 | THIS DATASET WILL BE ERROR FREE OR FREE OF HARMFUL COMPONENTS, ARE DISCLAIMED. 60 | IN NO EVENT SHALL THE COPYRIGHT HOLDER OR PREPARER OF DATASET OR ITS COMPANY 61 | OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 62 | EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 63 | OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 64 | INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 65 | CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 66 | IN ANY WAY OUT OF THE USE OF THIS DATASET EVEN IF ADVISED OF THE POSSIBILITY OF 67 | SUCH DAMAGE. 68 | 69 | You understand and agree that you will not: 1) make false or misleading 70 | statements or representations regarding Fujitsu or Fujitsu products and 71 | services; 2) take on any obligation or responsibility, or make any 72 | representation, warranty, guarantee or endorsement to anyone on Fujitsu's behalf 73 | (including, without limitation, any of our products or services); and that you 74 | will not state or imply that Fujitsu has developed, endorsed, reviewed or 75 | otherwise approved of any of your products particularly through the use of this 76 | DATASET. You agree that you will indemnify, defend and hold us harmless from 77 | and against any and all claims which may arise under or out of your use of the 78 | DATASET whether lawful or unlawful and you shall not be in violation of any of 79 | the terms of service or other terms of the github.com website or any other website 80 | where the DATASET is posted; your negligence or intentional misconduct; your 81 | products, or any integrations you develop, design, promote or distribute using 82 | the Software; any misrepresentations you make with respect to Fujitsu, or 83 | Fujitsu products or services. 84 | ``` 85 | --------------------------------------------------------------------------------