• Clair scanning fails with above error.
• /tmp on clair container host VM gets full.
• exact error message:

{"log": "

{\"level\":\"warn\",\"manifest\":\"sha256:6d2243864f75e2046008613ebb72ab92a3ebf92cb7608f9a264e18f35bd32c85\",\"state\":\"FetchLayers\",\"component\":\"internal/indexer/controller/Controller.Index\",\"error\":\"encountered error while fetching a layer: fetcher: unable to create file: open /tmp/fetch.2355163548: no space left on device\",\"time\":\"2022-08-02T09:50:32Z\",\"message\":\"layers fetch failure\"}

\n", "stream": "stderr", "time": "2022-08-02T09:50:32.386142917Z"}

$ grep -i "unable to create file: open" 0020-clair-log.txt | wc -l
16

• Similar/relevant bug : rpm: package scanner leaks extracted layers https://issues.redhat.com/browse/PROJQUAY-2315
• Qualcomm are running 4 instances of Clair and see 2 of them endup in with the disk filled in the /tmp partition. Is there a way to confirm the distribution across all Clair instances is working correctly?