Announcing FreeNAS 8.0.1-RELEASE

[louis-m]

i'd just like to say a big thank you to the devs etc. it's getting better all the time. thanks

 

[Milhouse]

I don't care too much about the nightly emails so I'll probably just wait for the official fix, but for now if you must receive the emails, edit the "daily_output" variable in /etc/defaults/periodic.conf and /conf/base/etc/defaults/periodic.conf to have your fully qualified email address and you should be back in business.

I quite appreciated the "security" log email, though I'd rather receive these emails ONLY when there is a problem (eg. failed security login) and not when nothing has happened. I'm really not sure what the purpose of the "daily run" email is, or why it has to be sent every day.

I guess an email filter may fix that, but an "email by exception" option would be appreciated, rather than spamming admins by default.

 

[ProtoSD]

I agree, an option for how often and what you want to be notified of would be nice. I opened a ticket for the email problem and William was quick to respond that it has been fixed and hinted that some kind of update should be available soon with fixes for that and some other things.

 

[Durkatlon]

It would be nice if you could receive an email only when something goes wrong, but that is not how those daily audit runs work unfortunately. Basically the audits get kicked off by running /usr/sbin/periodic with the "daily" argument. This script goes into /etc/periodic/daily and runs all the executable scripts in there. All the output from these scripts is collected in a temporary file. If this file is non-empty when ALL scripts are done, an email is sent. Of course quite a few of these scripts produce outputs even when everything is normal, so the email is always sent.

 

[ProtoSD]

Did anyone else try Durks fix for email? I tried it and got the same error/no email reports. (Changed BOTH files/locations)

 

[Durkatlon]

Did anyone else try Durks fix for email? I tried it and got the same error/no email reports. (Changed BOTH files/locations)

Hah, I didn't even try it myself. I guess I should :D. It seemed obvious that that would be the fix, but I guess I was wrong!

EDIT: OK, I *did* just try it and it worked. As root:

Code:

# vi /etc/defaults/periodic.conf
# cat /etc/defaults/periodic.conf | grep daily_output=
daily_output="myusername@myemaildomain.com" # user or /file
# /usr/sbin/periodic daily
#

Now wait a few minutes for the periodic script to finish and then you should get an email provided you didn't fat-finger the email address in the conf file. It worked perfectly on my end just now!

Note that the each type of run has it's own setting: "weekly_output", "montlhy_output", "daily_status_security_output".

 

[gcooper]

Hah, I didn't even try it myself. I guess I should :D. It seemed obvious that that would be the fix, but I guess I was wrong!

EDIT: OK, I *did* just try it and it worked. As root:

Code:

# vi /etc/defaults/periodic.conf
# cat /etc/defaults/periodic.conf | grep daily_output=
daily_output="myusername@myemaildomain.com" # user or /file
# /usr/sbin/periodic daily
#

Now wait a few minutes for the periodic script to finish and then you should get an email provided you didn't fat-finger the email address in the conf file. It worked perfectly on my end just now!

Note that the each type of run has it's own setting: "weekly_output", "montlhy_output", "daily_status_security_output".

Aliases and email alert reporting were both broken in 8.0.1 release with the new email stuff. They're fixed now on trunk and they have been backported to the 8.0.1 release branch, but we really need to roll an 8.0.1-p1 or something...

 

[Durkatlon]

So does that mean if I re-download 8.0.1 today, I will get a version where this is working properly?

 

[ProtoSD]

Sounds like you'd need to build it from source, but an updated release should be on the way....

 

[gcooper]

No. If you build it from _scratch_, yes.. but the problem is that our branches were a mess this release so tags/8.0.1-RELEASE isn't guaranteed to have everything in branches/8.0.1 .

We're in the process of ironing out an official release process, so the next major release will be better.

 

[Joshua Parker Ruehlig]

Also wanted to report that since switching to 8.0.1 release from a recent daily build stopped my UPS email notifications. UPS reporting works (can see alerts in /var/log/messages) but unplugging my UPS no long emails me, but it worked in the last few daily builds.

Also noticing the daily reports are gone, but that has already been said.

 

[Durkatlon]

Same root cause as the daily emails gone missing, which as you noted is a known issue. There's a fix for that already ported into the 8.0.1 Subversion branch. I think the plan is to come out with a maintenance release of 8.0.1. If you can't wait, protosd has compiled up a build from source. You can try sending him a PM for the location.

 

[Joshua Parker Ruehlig]

Same root cause as the daily emails gone missing, which as you noted is a known issue. There's a fix for that already ported into the 8.0.1 Subversion branch. I think the plan is to come out with a maintenance release of 8.0.1. If you can't wait, protosd has compiled up a build from source. You can try sending him a PM for the location.

I personally can wait as this isn't a major issue for me. Thanks for the options though

 

[coe]

I get this annoying alarm after webgui update from RC2:

WARNING: The WebGUI Address could not be bind to 192.168.11.9, using wildcard

Has this been reported yet?

 

[ProtoSD]

Yes, https://support.freenas.org/ticket/799

 

[Milhouse]

I'll look at this this weekend. This might affect customers.

This issue seems to have been resolved by the actions taken in ticket #860 (export/auto-import).

 

[coe]

Same root cause as the daily emails gone missing, which as you noted is a known issue. There's a fix for that already ported into the 8.0.1 Subversion branch. I think the plan is to come out with a maintenance release of 8.0.1. If you can't wait, protosd has compiled up a build from source. You can try sending him a PM for the location.

Any news regarding the maintenance release for 8.0.1? Rough ETA would be much appreciated if you have one.

 

[Tekkie]

Just tried the 8.0.1 release upgrade failed :( and /data/ 0 -rw-r--r-- 1 root www 0 Oct 11 19:24 upgrade.failed doesn't help me any further any ideas on how to debug the issue?

 

[ProtoSD]

Hey Tekkie, haven't seen you around for awhile, welcome back. How did you try to upgrade? CD or GUI? Look at the date(s) on the 'freenas-v1.db' files and grab the one with the older date so you have a backup of your settings. There might actually be one with .BAK on the end. You might have to do a clean install and then force an upgrade on the database, I'll have to look for the command to do that.

Edit:

This is what gets executed when you do a GUI upgrade:

/usr/bin/xz -cd /var/tmp/firmware/firmware.xz | sh /root/update && touch /data/need-update

Probably if you copy your database back to /data and touch /data/need-update and reboot that should upgrade the database.

 

[gaelic]

Wanna upgrade as soon as possible from 8.0.0-stable. Can anyone tell me if there will be any trouble (up to a possible data loss) when saving the config, updating via cd and reapplying the config? And just out of curiosity: why is a online upgrade from one stable to the next one not possible?

Thanks for the work on freenas,

C