[bisq-contrib] Reliability and monitoring of key nodes on the Bisq network
Mike Rosseel
mike.rosseel at gmail.com
Tue Dec 19 11:02:51 UTC 2017
Hi all,
As the guy who implemented most of the monitoring, I feel obliged to respond :)
First some perspective on the 'failing of decentralisation':
- the load has never been higher
- most of the things were never monitored before
If you look at the latest stats ( http://188.166.2.139:8080/status <http://188.166.2.139:8080/status> ), price nodes, bitcoin nodes and seednode uptime are pretty stable and have not been producing slack messages in the past few days.
For the BTC nodes I would even think that they’re more stable than the average public node we used before - but the public pool of nodes is much larger so there was more room for error before.
As for the scope of the monitoring, I based myself off of this issue I created: https://github.com/bisq-network/exchange/issues/1014 <https://github.com/bisq-network/exchange/issues/1014>
My only regret is that I concentrated on the price nodes and the btc nodes because they were showing issues, whereas the main culprits are now probably the seed nodes - which in my monitoring just used the old script to check that they’re running.
So I’m very glad that Manfred implemented the seednode connection java code - I do think it’s a waste that this was not done within the existing monitoring project.
Going forward:
- QoS specs are indeed very much needed, even more so if node operator compensation is tied to it
- I’m not implementing new monitoring until those specs are in place
- I’m available for Pricenode spec definition, either as owner or as sparring partner
regards,
Mike
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bisq.network/pipermail/bisq-contrib/attachments/20171219/8bf00c18/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: Message signed with OpenPGP
URL: <http://lists.bisq.network/pipermail/bisq-contrib/attachments/20171219/8bf00c18/attachment.sig>
More information about the bisq-contrib
mailing list