Techniques for assessing container images for vulnerabilities
Techniques are described for assessing container images for vulnerabilities without actually scanning the container images. A vulnerability assessment system (VAS) is described that is configured to perform vulnerabilities assessment for container images. The VAS is configured to perform the vulnerability assessment without scanning the container images. In certain embodiments, the VAS calculates a vulnerability score for the container image where the vulnerability score is indicative of a probability that the container image contains a vulnerability.
1. A method performed by one or more processors by executing a set of computer-readable instructions, the method comprising:
determining, based upon metadata associated with a container image to be processed for vulnerabilities, the container image including a hierarchy of a plurality of layers;
calculating a vulnerability score for each layer in the plurality of layers based upon scan results generated from scanning a plurality of container images for vulnerabilities at a given time;
computing a vulnerability score for the container image based upon the vulnerability scores calculated for the plurality of layers by accessing the vulnerability scores calculated for each of the plurality of layers included in the container image, and based on a weight assigned to a position of the layer in the hierarchy of the plurality of layers;
generating a vulnerability report for the container image, the vulnerability report including the vulnerability score for the container image; and
determining whether to scan the container image based on the vulnerability report.
2. The method of claim 1 , wherein the plurality of container images does not include the container image for which the vulnerability score is computed.
3. The method of claim 1 , wherein calculating the vulnerability score for each layer in the plurality of layers comprises:
for a first layer in the plurality of layers, calculating a vulnerability score for the first layer based upon (a) a number of container images in the plurality of container images that include the first layer, and (b) a number of container images in the plurality of container images including the first layer that were found to contain a vulnerability.
4. The method of claim 1 , wherein calculating the vulnerability score for each layer in the plurality of layers comprises:
for a first layer in the plurality of layers, calculating the vulnerability score for the first layer as a ratio of (a) a number of container images in the plurality of container images that include the first layer, and (b) a number of container images in the plurality of container images including the first layer that were found to contain a vulnerability.
5. The method of claim 1 , wherein the plurality of container images includes the container image for which the vulnerability score is computed.
6. The method of claim 1 , further comprising:
determining, based upon the vulnerability score computed for the container image, whether an action is to be performed for the container image.
7. The method of claim 6 , wherein determining whether the action is to be performed comprises:
identifying a preconfigured rule; and
determining whether the action is to be performed based upon the vulnerability score computed for the container image and the identified rule.
8. The method of claim 6 , wherein determining whether the action is to be performed comprises:
determining that the action is to be performed if the vulnerability score computed for the container image is greater than a threshold value.
9. The method of claim 6 , wherein the action comprises performing a vulnerability scan on the container image.
10. The method of claim 6 , wherein the action comprises generating an ordered list of a set of container images and associated vulnerability scores, the set of container images including the container image for which the vulnerability score is computed, wherein the set of container images are ordered in the ordered list based upon their associated vulnerability scores and the container images in the ordered list are scheduled for a vulnerability scan based upon their order in the ordered list starting from a highest ranked container image in the ordered list.
11. The method of claim 1 , further comprising identifying a hierarchy between the layers in the plurality of layers in the container image, wherein computing the vulnerability score for the container image comprises:
generating the vulnerability score for the container image based upon the vulnerability scores calculated for plurality of layers, and for each layer, a weight assigned to the layer based upon a memory size of the layer.
12. The method of claim 1 , further comprising identifying the hierarchy between the layers in the plurality of layers in the container image, wherein computing the vulnerability score for the container image comprises:
generating the vulnerability score for the container image based upon the vulnerability scores calculated for the plurality of layers in the hierarchy, and for each layer, the weight assigned to the layer based upon a most recent scan time of the layer.
13. The method of claim 1 , further comprising identifying a hierarchy between the layers in the plurality of layers in the container image, wherein computing the vulnerability score for the container image comprises determining an arithmetic mean of the vulnerability scores calculated for the plurality of layers.
14. The method of claim 1 wherein computing the vulnerability score for the container image comprises:
identifying, from a plurality of computation techniques, a first computation technique to be used for computing the vulnerability score for the container image; and
generating the vulnerability score for the container image by using the identified first computation technique.
15. The method of claim 1 , wherein computing the vulnerability score for the container image based upon the vulnerability scores calculated for the plurality of layers comprises:
generating a first vulnerability score for the container image based upon the vulnerability scores calculated for the plurality of layers;
determining a time period since a previous vulnerability scan of the container image;
generating a second vulnerability score for the container image based upon the first vulnerability score and the time period since the previous vulnerability scan of the container image, wherein the second vulnerability score is different from the first vulnerability score; and
outputting the second vulnerability score as the vulnerability score for the container image.
16. A system, comprising:
one or more processors; and
a non-transitory computer-readable storage medium comprising computer-executable instructions that, when executed by the processor, cause the system to perform processing comprising:
determining, based upon metadata associated with a container image to be processed for vulnerabilities, the container image including a hierarchy of a plurality of layers;
calculating a vulnerability score for each layer in the plurality of layers based upon scan results generated from scanning a plurality of container images for vulnerabilities at a given time;
computing a vulnerability score for the container image based upon the vulnerability scores calculated for the plurality of layers by accessing the vulnerability scores calculated for each of the plurality of layers included in the container image, and based on a weight assigned to a position of the layer in the hierarchy of the plurality of layers;
generating a vulnerability report for the container image, the vulnerability report including the vulnerability score for the container image; and
determining whether to scan the container image based on the vulnerability report.
17. The system of claim 16 , wherein calculating the vulnerability score for each layer in the plurality of layers comprises:
for a first layer in the plurality of layers, calculating the vulnerability score for the first layer based upon (a) a number of container images in the plurality of container images that include the first layer, and (b) a number of container images in the plurality of container images including the first layer that were found to contain a vulnerability.
18. A non-transitory computer-readable storage medium comprising computer-executable instructions that when executed by a processor, cause the processor to perform processing comprising:
determining, based upon metadata associated with a container image to be processed for vulnerabilities, the container image including a hierarchy of a plurality of layers;
calculating a vulnerability score for each layer in the plurality of layers based upon scan results generated from scanning a plurality of container images for vulnerabilities at a given time;
computing a vulnerability score for the container image based upon the vulnerability scores calculated for the plurality of layers by accessing the vulnerability scores calculated for each of the plurality of layers included in the container image, and based on a weight assigned to a position of the layer in the hierarchy of the plurality of layers;
generating a vulnerability report for the container image, the vulnerability report including the vulnerability score for the container image; and
determining whether to scan the container image based on the vulnerability report.
19. The method according to claim 1 , wherein a higher weight is given to a higher layer, and a lower weight is given to a lower layer.
20. The method according to claim 1 , wherein a higher weight is given to a layer with a larger memory size, and a lower weight is given to a layer with a lower memory size.