FWD » Core Development » Runtime Infrastructure

From Igor:

I've not seen this, but, as far as I remember, in such situation a new SSL handshake will be initiated, so it is not a fatal error.

If this is non-fatal, then perhaps we should reduce the logging level so that it is only seen when there is a more detailed logging enabled. We don't need to create the impression that there are problems under normal circumstances when there really is no issue.

I observed the case when the server threw

[05/26/2021 19:40:41 MSK] (com.goldencode.p2j.net.SSL:WARNING) handshake failure
javax.net.ssl.SSLException: bad record MAC
        at sun.security.ssl.Alerts.getSSLException(Alerts.java:214)
        at sun.security.ssl.SSLEngineImpl.fatal(SSLEngineImpl.java:1729)
        at sun.security.ssl.SSLEngineImpl.readRecord(SSLEngineImpl.java:987)
        at sun.security.ssl.SSLEngineImpl.readNetRecord(SSLEngineImpl.java:913)
        at sun.security.ssl.SSLEngineImpl.unwrap(SSLEngineImpl.java:783)
        at javax.net.ssl.SSLEngine.unwrap(SSLEngine.java:626)
        at com.goldencode.p2j.net.SSL.unwrap(SSL.java:484)
        at com.goldencode.p2j.net.SSL.handshake(SSL.java:386)
        at com.goldencode.p2j.net.SSL.run(SSL.java:256)
        at com.goldencode.p2j.net.SSL.notify(SSL.java:247)
        at com.goldencode.p2j.net.BlockingSSL.processInput(BlockingSSL.java:207)
        at com.goldencode.p2j.net.BlockingSSL.checkInput(BlockingSSL.java:113)
        at com.goldencode.p2j.net.NIOSslSocket.lambda$new$1(NIOSslSocket.java:172)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
        at java.lang.Thread.run(Thread.java:748)
Caused by: javax.crypto.BadPaddingException: bad record MAC
        at sun.security.ssl.EngineInputRecord.decrypt(EngineInputRecord.java:238)
        at sun.security.ssl.SSLEngineImpl.readRecord(SSLEngineImpl.java:980)
        at sun.security.ssl.SSLEngineImpl.readNetRecord(SSLEngineImpl.java:913)
        at sun.security.ssl.SSLEngineImpl.unwrap(SSLEngineImpl.java:783)
        at javax.net.ssl.SSLEngine.unwrap(SSLEngine.java:626)
        at com.goldencode.p2j.net.SSL.unwrap(SSL.java:484)
        at com.goldencode.p2j.net.SSL.handshake(SSL.java:386)
        at com.goldencode.p2j.net.SSL.run(SSL.java:256)
        at com.goldencode.p2j.net.SSL.notify(SSL.java:247)
        at com.goldencode.p2j.net.BlockingSSL.processInput(BlockingSSL.java:207)
        at com.goldencode.p2j.net.BlockingSSL.checkInput(BlockingSSL.java:113)
        at com.goldencode.p2j.net.NIOSslSocket.lambda$new$1(NIOSslSocket.java:172)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
        at java.lang.Thread.run(Thread.java:748)

To clarify #5362-5 I added messages logs for the js web client

websocket is opened. p2j.socket.js:4624:18
1: MSG_CREATE_WINDOW of length 6 done in 24 p2j.socket.js:4376:18
2: CAPTURE_MOUSE of length 2 done in 0 p2j.socket.js:4376:18
3: MSG_IS_FONT_INSTALLED of length 50 done in 2 p2j.socket.js:4376:18
4: MSG_CREATE_FONT of length 58 done in 16 p2j.socket.js:4376:18
5: MSG_DERIVE_FONT of length 9 done in 0 p2j.socket.js:4376:18
6: MSG_IS_FONT_INSTALLED of length 32 done in 1 p2j.socket.js:4376:18
7: MSG_CREATE_FONT of length 40 done in 15 p2j.socket.js:4376:18
8: MSG_DERIVE_FONT of length 9 done in 1 p2j.socket.js:4376:18
9: MSG_IS_FONT_INSTALLED of length 26 done in 4 p2j.socket.js:4376:18
10: MSG_IS_FONT_INSTALLED of length 28 done in 2 p2j.socket.js:4376:18
11: MSG_CREATE_FONT of length 217396 done in 24 p2j.socket.js:4376:18
downloadable font: kern: Too large subtable (font-family: "Open Sans" style:normal weight:400 stretch:100 src index:0) source: data:application/x-font-truetype;base64,AAEAAAATAQAABAAwRFNJR54SRB0AAzucAAAVdEdERUYAJgOvAAM3fAAAAB5HUE9TCzcPNwADN5wAAAA4R1NVQg4r … k2BOG8QrV5WUyxbhmmcz0SspE8jsvrxZsQOkKZnsHYiZJId3D5vKFPvU1ElMdA7IPS5vIMkDzejlD9AhObNWGdX7rL2sqTi9sNUMo9ljrbCVtGhYw+LXKf+RpMcAAA==
downloadable font: Table discarded (font-family: "Open Sans" style:normal weight:400 stretch:100 src index:0) source: data:application/x-font-truetype;base64,AAEAAAATAQAABAAwRFNJR54SRB0AAzucAAAVdEdERUYAJgOvAAM3fAAAAB5HUE9TCzcPNwADN5wAAAA4R1NVQg4r … k2BOG8QrV5WUyxbhmmcz0SspE8jsvrxZsQOkKZnsHYiZJId3D5vKFPvU1ElMdA7IPS5vIMkDzejlD9AhObNWGdX7rL2sqTi9sNUMo9ljrbCVtGhYw+LXKf+RpMcAAA==
12: MSG_DERIVE_FONT of length 9 done in 0 p2j.socket.js:4376:18
13: MSG_IS_FONT_INSTALLED of length 26 done in 2 p2j.socket.js:4376:18
14: MSG_IS_FONT_INSTALLED of length 28 done in 1 p2j.socket.js:4376:18
15: MSG_CREATE_FONT of length 36 done in 16 p2j.socket.js:4376:18
16: MSG_DERIVE_FONT of length 9 done in 0 p2j.socket.js:4376:18
17: MSG_IS_FONT_INSTALLED of length 26 done in 4 p2j.socket.js:4376:18
18: MSG_IS_FONT_INSTALLED of length 28 done in 0 p2j.socket.js:4376:18
19: MSG_CREATE_FONT of length 36 done in 12 p2j.socket.js:4376:18
20: MSG_DERIVE_FONT of length 9 done in 0 p2j.socket.js:4376:18
21: MSG_IS_FONT_INSTALLED of length 26 done in 1 p2j.socket.js:4376:18
22: MSG_IS_FONT_INSTALLED of length 28 done in 2 p2j.socket.js:4376:18
23: MSG_CREATE_FONT of length 36 done in 18 p2j.socket.js:4376:18
24: MSG_DERIVE_FONT of length 9 done in 0 p2j.socket.js:4376:18
25: MSG_IS_FONT_INSTALLED of length 22 done in 3 p2j.socket.js:4376:18
26: MSG_CREATE_FONT of length 30 done in 14 p2j.socket.js:4376:18
27: MSG_DERIVE_FONT of length 9 done in 0 p2j.socket.js:4376:18
28: MSG_SET_UPLOAD_FILE_SIZE_LIMITS of length 9 done in 0 p2j.socket.js:4376:18
30: MSG_CLIENT_READY of length 1 done in 1 p2j.socket.js:4376:18
31: MSG_GET_DESKTOP_DIMENSION of length 5 done in 0

If we cannot recover the session, then we need to exit the web client. BUT if possible, we should try to reconnect the websocket without exiting the web client.

Igor: Do you think this is this resolved by your changes in #5388?

Greg Shah wrote:

Igor: Do you think this is this resolved by your changes in #5388?

Greg,
Yes, I believe that it was the same problem.