Christoph John

Would be good if this could also be merged to 2.2.x then. It causes intermittent failures in our test suite (QuickFIX/J)

@harinda05 Thanks for the PR. Could you please base this against master? Thanks in advance

Seems to be related to #399 Any idea how this could be reproduced without two nodes since I don't have such a setup? Could you please share your session settings?

Are you able to reproduce this in a test environment? If yes, could you test a pull request branch in your test environment? The only solution I could come up...

Currently the `IOException` causes the message store to never close the old store and open new files for storing. This error then happens on every access to the store, as...

> Generally, in case of IOException there should be a finally block which must close this files or this is something which is currently missing Currently the `FileStore` tries to...

> Otherwise, we will have to somehow handle the exception to ignore IOException. Do you have any idea in that case what can be done? I am not sure, but...

At some point (depending on how much messages you will send/receive) the sequence number will probably overflow (AFAIR we have a bug there). Moreover, the larger your store gets, the...

Hi @leonchen83 @philipwhiuk et al so it seems we have two options if we don't want time to go backwards (related to time of the previous message): Either store the...

> As the latter is already in use, perhaps it would be time to disable LGTM on this repository. Thanks for the reminder. I've disabled it now. Hopefully the check...