Hi. Can the amount of memory in an SN have an affect on the bandwidth requirements? Logic would think it wouldn’t as bandwidth is more a network related concern. I have had some Computing Power failures in the past, and was planning increasing memory.
Over the last tow days been getting bandwidth failures, not sure why as have been passing the tests for over a year.
As a side note: noticing some chain part and height failures on some of the Alices.
Low memory can cause code getting swapped out and then moving the code into ram again might be a problem. Also when low on memory it is more likely to have a garbage collection when the test is done which can slow down the entire node.
I already restarted the 5 nodes that had problems, thx for mentioning.
Thanks BR. Will sort this out. Much appreciated as always.
Thank you very much. Now it is clearer for me :).
Hi, I’ve just changed the IP of my SN. Changed it in the appropriate servant and NIS config files. The only remaining thing I need to do is send the change IP message on here. However, before doing that I need to change the delegated harvesting IP on the NCC, but when trying to logon to the client, it comes back with NIS not availabled (probably because it’s referring to the old IP in the client). Is there a way to logon to my wallet, or manually edit a config file on the Windows client (v x.x.93) to point to the new delegated IP address - thus allowing me to logon?
You can change the configuration (e.g. the host) in the landing page.
Sorry, found the post at the top. Will send a message to the nafund bukios account, then let you know.
What would be the optimal setting for the servantStart.sh file - 128 or 256? I’m not familiar with these settings, are they reserving memory for the service? i.e. if I changed it from 256 to 128 then I would have more free memory available for other operations on the SN?
Thanks BloodyRookie. Appreciated, I didn’t know that was possible.
Usually 128MB for the servant are enough. In rare cases when the servant has connection problems, it can run out of memory with 128MB.
If you decrease from 256MB to 128MB, then you have indeed 128MB more memory for NIS.
But if you experience memory problems, in general it is better to rent a server with more memory, servers are not that expensive.
Thanks BR, I did upgrade to more memory this morning, but it was still showing as almost 2GB of memory being used. BTW, did you get my change IP message?
Which supernode was yours?
If a SN owner increases the number of nodes allowed to connect to their SN for harvesting to 20 or 50 for example would it have an impact on the performance of their server or ability to harvest blocks?
No, it should neither affect the performance nor the ability to harvest a block.
Why do we limit it to 4(?) connections only by default?
non-supernodes probably don’t want tons of harvesters on their server.
Hello, how long does it take to add supernodes to the listing?
should not take more than 15 mins. I guess you enrolled a supernode, can you post the tx hash for the enroll?
My enroll txHash is 851b4049282fa1d287bcbc2fa0dffd98a6a85e9eac11206c8b1f9d9bd9f49f4e. Maybe i did something wrong. Thanks!