NEM Supernode Rewards Program

Blog Posts
#1923

http://chain.nem.ninja/#/transfer/be19769bb94a31eb39d3a6655fdff8ef3f5c21e4b9808087b7d6e0625f20a153

The day before it didn’t receive a payout cause it had failed a test

#1924

Thank you for your reply.
change ip 183.181.8.127 to dawon.supernode.me
is the right message?

sample message was filled with 0 if the ip address contains two figures.
anyway , I guide to re-send above message, thank you again.

#1925

Database maintenance finished, master is started again.

2 Likes
#1926

@BloodyRookie
Thank you for your hard work.

2 Likes
#1927

@BloodyRookie
Thank you for your work!

and I have one question.
Today supernode reward for Dec/17/2017 seemed not to be sent.

Please check if the reward system works fine again.

Thanks

#1928

Due to the maintenance the payout is delayed, please wait a bit more.

1 Like
#1929

Thank you for your reply and your hard work again.

No warry, safety first please :slight_smile:

#1930

If height, chain part, version and responsiveness are failing, but the other test are passing, does that mean NIS has crashed?

#1931

When using NIS servers for a long time, Java garbage collector errors occur.
This problem will be resolved by restarting the NIS server.
From now on, we recommend that you restart the NIS server once a month.

thanks

#1932

What do you mean with “Java garbage collector errors occur”?

1 Like
#1933

To be exact, it is “GC overhead limit exceeded”.

#1934

Probably memory gets too fragmented over time when you only give 1 GB to NIS.
The more memory you can give to NIS the better it is. HugeAlice is running for 140 days without problems but i gave 16GB to NIS ^^
(and use G1 garbage collector to minimize the stop the world events)

1 Like
#1935

That is probably a good idea.

1 Like
#1936

For general people, I think that 1 GB allocation is the most current.
With this setting I had a problem in 3 months.
From a preventive point of view, we believe that nearly everyone will not get this error when restarting once a month.
(In an environment where multiple threads can be used, I think G1GC is a good solution.)

#1937

Configuration

Windows Server 2012 R2 Datacenter

Plan: G2.4GB

CPU: 2 vCPU

RAM: 4 GB

Disk: 100 GB

Term: On-Demand

Transfer: 4096 GB
This configuration works stably …
2 gigabytes of RAM is small.

#1938

This configuration is for supernodes.

#1939

4GB is better than 2GB, that is true, but 2GB does work for a while. I have 2GB nodes and just need to restart NIS once in a while.

#1940

I’m using 3GB for my supernode, but it had the “garbage collector” problem after 2 weeks running.
I’m going to restart it every week now.

#1941

3GB for NIS?

#1942

3GB in total, I use standard settings for NIS memory