[bisq-network/bisq] High system load on Arch linux (#4003)

[Florian Reimair]

stejbac:
> I don't think #4025 would affect the hotspot in the screenshot, as it makes no important change to the way SequenceNumberMap is serialised from FileManager. I think that class (like a number of others implementing PersistableEnvelope) already had thread-safe proto conversion, as it just had a single ConcurrentHashMap field. So I didn't force it to run in the user thread. The majority of classes implementing PersistableEnvelope were made to do proto conversion (during persistence) in the user thread, but I'm fairly sure that wouldn't help performance - it might make it harder to spot in the JProfiler views though. It looks like the JProfiler run in the screenshot was in full instrumentation mode - I'm not sure it that might be affecting the apparent hotspot at all, but when I profiled the same area of the code in sampling mode the hotspot seemed to be coming from the sync method call rather than proto conversion.

-- 
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/4003#issuecomment-608344825
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bisq.network/pipermail/bisq-github/attachments/20200403/124cab2c/attachment.html>