Sogo 3.x.x released in nethforge-testing

transocean · February 10, 2017, 8:36am

Moin Ralf,

I have tried here Nethserver 7.3 Final with SOGo 3.2.6a as VM and also on the described real hardware. The described behavior is the same for both variants.

Regards

Uwe

flatspin · February 10, 2017, 9:12am

Hi Uwe,
Sorry, maybe my post was unclear. My VM is also a NS7-final installation migrated from RC4.1 with SOGo 3.2.6a. I use virtualbox on the TX100S3 with NS6.8 installed.
What migration path do you have?

transocean · February 10, 2017, 9:19am

Hi Ralf,

i have the same Migration path. I will test now with a fresh Final on my Hyper V 2008 R2.

stephdl · February 16, 2017, 9:37pm

Sogo released in nethforge, you should not need anymore to restart the sogod service, the process is automatic.

Stefano_Zamboni · February 16, 2017, 9:49pm

sogo 3.2.6a on a vm with 10 GB of ram and 220 users… no issue… please be sure not to have squid/proxy enabled

stephdl · February 17, 2017, 12:00am

sogo-3.2.7 in nethforge-testing

for a first install

yum install nethserver-sogo --enablerepo=nethforge-testing

for an upgrade

yum update sogo\* sope\* --enablerepo=nethforge-testing

alefattorini · February 17, 2017, 9:36am

That’s a great news. Ehi @transocean @quality_team @hucky @mark_nl @asl @frx44 could you help that guy?

transocean · February 17, 2017, 6:47pm

No problem. I will test it tomorrow with a fresh installation of NS7 Final on my Hyper-V Server.

Regards

Uwe

hucky · February 18, 2017, 4:37am

have update it yesterday and it seems ok

transocean · February 18, 2017, 9:53am

Hi @ all,

I installed NS 7 as a virtual machine with four virtual processors and 8 GB RAM. As an account provider, I use LDAP. SOGo in version 3.2.7 runs on the virtual machine as a brand new installation. Everything looks good and seems to work well. But I have as before a problem with the speed of SOGo. In my opinion it works too slowly.
When switching between calendar and mail, it sometimes takes up to five seconds for the system to respond. This did not exist in the earlier SOGo version (2 …) on a NS 6.8 machine.

I have NS7 final installed on a physical machine. AccountProvider is LDAP. Mail done Roundcube. And for sharing, Nextcloud works. The system is in the DMZ and does a good job. Originally, I had planned to replace Roundcube with SOGo. At the moment I do not want to do this because SOGo is too slow for me.

Here in the forum, a user has expressed the idea to use HORDE Groupware instead of SOGo. A suggestion that I like very much. I had before Nethserver beside my Windows servers also always a SME server, which has worked with HORDE Groupware. This was a very reliable matter. Unfortunately I miss the Linux experience. Otherwise, I would like to try as a test system.

Regards and a nice weekend…

Uwe

stephdl · February 18, 2017, 10:18am

I suppose a brand new with nothing at all ?

I can reproduce it but my machine gets some rpm installed (java with madsonic, samba AD…)

Stefano_Zamboni · February 18, 2017, 10:19am

it’s a quite complex web app… 5 seconds are acceptable… there are webmail that are slower out there

and you can’t compare sogo2 with sogo3, they just share the name

transocean · February 18, 2017, 10:28am

I suppose a brand new with nothing at all ?

Yes.

transocean · February 18, 2017, 10:29am

I now…

stephdl · February 18, 2017, 10:46am

for a matter of tests, can you play with prefork value in

vim /etc/sysconfig/sogo

increase it, about 15, 20, 30…

restart the service

systemctl restart sogod

I have the feeling that it is better :-?

stephdl · February 18, 2017, 10:51am

in fact the good command is

config setprop sogod WOWorkersCount 30
signal-event nethserver-sogo-update

transocean · February 18, 2017, 10:52am

Okay,

i will test it later. Just now my wife is calling. I have to do some work here in the house.

stephdl · February 18, 2017, 10:54am

good luck, my wife is behind my back today

gerald_FS · February 18, 2017, 10:59am

Have fun - women love shopping

stephdl · February 18, 2017, 11:24am

It happend, because you didn’t have enough workers for the rush.
For such rush traffic you also could increase WOListenQueueSize.
This tells sogod to queue up that count of requests for its workers.
This will not decrease your workload, but clients are not rejected
instantly, when no worker is available.

You do not need to kill those workers by yourself.
Instead you can decrease WOWatchDogRequestTimeout (in minutes) which is
set to 10 minutes by default.
Then the SOGo WatchDog will kill those processes for you.

Your RAM should be approx.:
worker count * SxVMemLimit (which is in MB)

In your case 20 * 384 MB = 7680 MB (~ 8GB)

With 20 workers you should be able to handle all 200 users, as long as
they use web interface and/or calDAV/cardDAV.

If you have ActiveSync users, you have to reserve one worker per
ActiveSync connection!

Kind regards,
Christian Mack

an interesting link https://lists.inverse.ca/sogo/arc/users/2015-09/msg00196.html