Potential issue with submitted transactions?

[mikera]

Need to check code handling new block production in cases of faulty / missing transactions

Should not cause any lockup, should result in immediate error to client

Hi @Mike - the MissingDataException re-occurred about 3hrs after a restart. I've attached the log file. But I know that this alone isn't enough to diagnose the problem (probably various resource stats also needed). But on the slim chance that it does remind you of something, please let me know. 
Loading key store at: /home/am215/.convex/keystore.pfx
16:47:17.762 [main] INFO  convex.peer.Server - Restored Peer with root data hash: 0x549c648e817897594253056ccc184a5309ea4a52b4e367ba62e29f6876381ad2
16:47:17.928 [main] INFO  convex.peer.Server - Peer server started on port 18888 with peer key: 0x8c5c69c889e954491fa1cd490742e2d4615fd495e10ae89403f07a02decb6e9d
java.nio.file.NoSuchFileException: /home/am215/.convex/ssl/certificate.pem
	at java.base/sun.nio.fs.UnixException.translateToIOException(UnixException.java:92)
	at java.base/sun.nio.fs.UnixException.rethrowAsIOException(UnixException.java:106)
Expand
convex-truncated-20251113.log
16 KB
[09:27]Mike: Thanks @AshMcClenaghan it's not quite enough to identify the cause, but it is definitely happening in the code where your peer tries to produce a new block, so was probably triggered by something like a faulty transaction? Did you submit a transaction at the tim eit happened?
[09:28]Mike: That error could be caused by a bit flip error on the machine (e.g. changing a bit in hash so the lookup fails). That should be rare, but it is a possibility....
[09:46]AshMcClenaghan: Thanks for taking the look at it @Mike. I've 25 students plus an automated process all firing txs at it so... a faulty tx... possibly but should that result in the receiving peer effectively locking up(?).  (If I'd time, I'd build my dApp against your latest commit but for now its running against an old one from 20th Aug.)
[09:50]Mike: Yes it could be a bug, possibly in the transaction receiving handling code. I'll see if I can add some error handling that might prevent a lockup (invalid transactions should in theory always fail and get rejected)