[bisq-network/bisq] bisq v1.6.5 hanged on shutdown attemp (#5561)

Stan notifications at github.com
Thu Jun 10 17:23:32 CEST 2021


This problem may be related to one that was blocking work on atomic-tx, and fixed in https://github.com/bisq-network/bisq/pull/5528 (but not in v1.6.5).  
Raising the MAXRam limit might not solve this.

The problem may be in https://github.com/bisq-network/bisq/blob/release/v1.6.5/core/src/main/java/bisq/core/provider/fee/FeeService.java LINES 140-144

	public void requestFees(@Nullable Runnable resultHandler, @Nullable FaultHandler faultHandler) {
		if (feeProvider.getHttpClient().hasPendingRequest()) {
			log.warn("We have a pending request open. We ignore that request. httpClient {}", feeProvider.getHttpClient());
			                       // Should be: "... We ignore this new request ..."
			return;
		}
		...
	}       
	 
New fee requests are ignored and some pending request callbacks never complete;  this may be why shutdown is hanging.  

@sqrrm's PR may solve this problem too.  It adds a new callback for each new request if there is already pending request, rather than just ignoring the new request and returning.  Then, on success or failure of any callback, all other pending callbacks' respective resultHandlers or faultHandlers are run.  This unwinding of callbacks' handlers might prevent hanging shutdown, but not sure about the OOME.  @sqrrm, is this an incorrect interpretation of https://github.com/bisq-network/bisq/pull/5528?

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/bisq-network/bisq/issues/5561#issuecomment-858716480
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bisq.network/pipermail/bisq-github/attachments/20210610/ee085676/attachment.htm>


More information about the bisq-github mailing list