Has anyone seen the condition when threads in an Atlassian app spin this way (eventually consuming all available CPU and/or thread pool)? If so, have you resolved it?
Doesn't seem to matter much where the TA library is called from; a good combination is JIRA + FishEye plugin.
java.lang.Thread.State: RUNNABLE
at java.math.BigInteger.mulAdd(BigInteger.java:1930)
at java.math.BigInteger.montReduce(BigInteger.java:1880)
at java.math.BigInteger.oddModPow(BigInteger.java:1850)
at java.math.BigInteger.modPow(BigInteger.java:1599)
at org.bouncycastle.crypto.engines.RSACoreEngine.processBlock(Unknown Source)
at org.bouncycastle.crypto.engines.RSABlindedEngine.processBlock(Unknown Source)
at org.bouncycastle.jce.provider.JCERSACipher.engineDoFinal(Unknown Source)
at javax.crypto.Cipher.doFinal(Cipher.java:2087)
at com.atlassian.security.auth.trustedapps.BouncyCastleEncryptionProvider.createEncryptedCertificate(BouncyCastleEncryptionProvider.java:229)
at com.atlassian.security.auth.trustedapps.DefaultCurrentApplication.encode(DefaultCurrentApplication.java:54)
at com.atlassian.applinks.core.auth.trusted.TrustedRequest.signRequest(TrustedRequest.java:67)
at com.atlassian.applinks.core.auth.trusted.TrustedRequest.signRequest(TrustedRequest.java:58)
at com.atlassian.applinks.core.auth.trusted.TrustedRequest.execute(TrustedRequest.java:38)
at com.atlassian.jirafisheyeplugin.rest.FishEyeRestApiManagerImpl.callFisheye(FishEyeRestApiManagerImpl.java:175)
<...>