Available Space and SMR HDD

[hafdan]

Hello together

I’ve been at 5TB for a while now and in the LOG there’s always this nightly news: “Available Space”: 4016706900480} is it possible to adjust that the system knows it has more space? Or just wait until it goes to 0?

Thanks, greetings Daniel

PS: I found out that I have WB SMR HDDs and added “filestore.write-buffer-size: 4MiB” to my config. So far everything OK!

[Vadim]

@hafdan

“filestore.write-buffer-size: 4MiB”

What exactly makes this command?

[Alexey]

Filestore.write-buffer-size

Hi guys, I’m sorry I keep asking questions. I was going through the config.yam and came across the # in-memory buffer for uploads option. I did search the forums but didn’t find any information regarding this option. Is it worth enabling something like 512MiB or even 1GiB as write cache? Will it improve node performance? I understand it does increase the risk of data corruption in case of an ungraceful shutdown but I’m behind a UPS so hopefully that is somewhat unlikely. Any thoughts? André …

I can only found this feedback:

Machine freezing for short periods

After roughly 1h15 with the --filestore.write-buffer-size="4MiB" option, the node is dying again it seems: load average: 299.51, 282.63, 210.63. Besides, something really suspect happens time to time, not sure if it was the case with previous configurations, but it is little by little taking a very large amount of RAM (1000MB+) up to the point where it suddenly goes back to almost nothing, and it rises again (I saw it above 1580MB once). When checking the logs, it appears that the start-up rou…

So this option is likely doesn’t help.

[Alexey]

The storagenode puts to the log the number reported by your OS.

[r51f]

This is what you need to set in storj config.yaml.
As I can understand from https://github.com/storj/storj/issues/3854, this changes in-memory write buffer to get better perfomance for SMR drives.