“1. Pick an external storage, which you want to move your persistent data to.
6. …your data will from now on be written to the external disk.”
What do you mean by “your data will from now on be written to the external disk”? If it means just moving of /srv, I would welcome if it was written in Foris in a clearer way. I.e. what you will move to a new location for us and what we have to (or should) move manually after that. You write about “Nextcloud, LXC or other IO intensive applications” but it seems that Storage plugin will move /srv only. That doc page seems to me a little bit confusing.
Your description is clear to me. It means that it does nothing with our persistent data in general, it will only move /srv to a new device/partition and gives us possibility to move data from other applications to that location after that. But we need to know what they are and where they are and move them manually. This plugin takes care about LXC only.
Someone would think his internal storage is safe after this automatic move but if he had IO intensive application outside LXC, it is not the case.
Please change description in Foris and the Doc page to make it clear. Thank you.
This is just a first RC version, which we released. We have the issue for it on our Gitlab. Right now we want to fix it to 3.10, but something can change.
I moved /srv (storage) and can’t reach FORIS now. Neither luci or ssh…
I have external drive with three partition (sdc1, sdc2, sdc3). I choose sdc3 and before reboot I saw that disc was formated. Originaly I had btrfs, but after save I saw it was formated to ext4.
LAN working with internet connection.
Wifi working with internet connection.
LXC containers don’t work (there are on sdc1).
I saved configuration with new tool so tomorrow till probably make factory reset and new update to RC with downloading backup of settings
We’d like to reproduce it, but we need to see what happened there.
Can you attach serial link to your router and send us necessary log as /var/log/messages and output of these commands dmesg, lsblk and opkg list-installed to email tech.support@turris.cz?
I sent output of commands to email. /var/log is almost empty.
Today I noticed that wifi is not working aswell, but LAN is OK. Therefore I connected my wife to LAN
As we are going out for extended weekend I will make reset at Monday evening. Till then you can check messages and I can test something before reset if needed.
I just run update today. Everything seems to work ok, but I see this errors during update:
Output from foris-netmetr-plugin.postinst:
[ -n ]
/etc/init.d/lighttpd restart
Output from pakon-lists.postinst:
Traceback (most recent call last):
File “/usr/libexec/pakon-light/domains_reapply.py”, line 75, in
main()
File “/usr/libexec/pakon-light/domains_reapply.py”, line 71, in main
replace(’/var/lib/pakon.db’, multiple_replace)
File “/usr/libexec/pakon-light/domains_reapply.py”, line 55, in replace
for row in c.execute(‘SELECT DISTINCT(app_hostname) FROM traffic WHERE app_hostname IS NOT NULL’):
sqlite3.OperationalError: no such table: traffic
Command failed: Not found
Command failed: Not found
Output from suricata.prerm:
Command failed: Not found
Command failed: Not found
We just released an update, it contains some more polishing of storage plugin, fix to Suricata dualstack DNS issue and few other small fixes and updates.
it would be great if you updated https_dns_proxy to 2018-04-23. Turris has an ancient version, the current version supports Cloudflare’s 1.1.1.1 DNS, which is faster than Google’s and using DNS over HTTPS would be pretty cool as a privacy feature.
Whilst certainly appreciating the testing/refining of this RC I am wondering what is the point of tying up the release (and thus holding back) of updated packages like knot-resolver, unbound, netdata etc, all of which can be configured without an UI and via sshd instead?
