Seafile Pro 6.2 is ready!

Thank you. I will looking for 6.3 release.

Thanks for clarification but then I wonder why I got a different answer in March this year by email: ‘Our latest version support PostgreSQL too. We only don’t provide upgrade and setup scripts for PostgreSQL.’
The topic was also about Pro, not CE back then.

One of my threads six months ago was also about the Pro version which was worked on and fixed.

I upgraded from pro 6.1.8 to pro 6.2.4. I am using a seafile cluster (two applications servers, one background node) in combination with a real-time backup.

the upgrade worked almost perfectly. the only issue I see is, that the real-time backup server cannot load the repo-list anymore:

http-tx-mgr.c(514): libcurl failed to GET https://[primary server]/seafhttp/server-sync/repo-list: Problem with the SSL CA cert (path? access rights?).

My systems are running debian and I am aware of this:
https://manual.seafile.com/deploy_pro/real_time_backup.html

specifically the part on bundling the ca-certificates.

did something change here? before the upgrade, the real-time backup was running fine (with the same server certificates).

or could this be caused by the switching from fastcgi to WSGI? (although as far as I understand this affects only the seahub service).

any hints aprreciated.

Best,
Hp

This maybe caused by that the new version is packaged in a CentOS 7 machine, leading to some incompatibility with the system libraries.

Mark,
Same as your problem after upgrade from 6.1.9 pro to 6.2.4 pro edition.Difference is the SSL fail with S3 storage backend.
My seafile.log said:
[12/17/17 12:51:28] …/common/s3-client.c(552): [S3] libcurl failed to get bucket-fs:df0d4830927d053f6e762da97d2466091fb894dc: Problem with the SSL CA cert (path? access rights?).
[12/17/17 12:51:28] …/common/s3-client.c(552): [S3] libcurl failed to get bucket-fs:df0d4830927d053f6e762da97d2466091fb894dc: Problem with the SSL CA cert (path? access rights?).
[12/17/17 12:51:28] …/common/s3-client.c(552): [S3] libcurl failed to get bucket-fs:df0d4830927d053f6e762da97d2466091fb894dc: Problem with the SSL CA cert (path? access rights?).
[12/17/17 12:51:28] …/common/fs-mgr.c(1893): [fs mgr] Failed to read dir df0d4830927d053f6e762da97d2466091fb894dc.
Now I’m also using Debian 9,and also switched from fastcgi to WSGI.
Mark again!

1 Like

Mark.
Same problem with hkunz.
System CA bundle issue after upgrade from 6.1.9 pro to 6.2.4 pro.
Despite once worked well with
https://manual.seafile.com/deploy_pro/setup_with_amazon_s3.html
Mark again.

good to know, thanks. is there some workaround?
I am using jessie right know, but will soon move to stretch. could that solve the issue?

best,
hp

Hi,
I upgraded our Seafile server to version 6.2.4 recently and decided to tune the Seahub login page.
I uploaded a background image and a logo, according to the specified image sizes.
My login page now looks great, but my logo is all squished in all the other admin pages…
Could you please use the same logo ration in all pages, please ?
I can send you screenshots if you need…
Thanks !

1 Like

Had the same problem, deleting

LOGO_PATH
LOGO_WIDTH
LOGO_HEIGHT

from seahub_settings.py resolved it, maybe that helps you too.

1 Like

Hello Simsala,

You advice “did the trick” :+1:

Thank you very much for your answer and have a nice day !

1 Like

What version for Seafile PRO is better. We using latest Debian stable release so I think the Seafile PRO ubuntu version is what I need cause ubuntu is build on debian, but want be sure.

There is “classic” and ubuntu tagged version in customer portal so that’s why am I asking.

The ubuntu version is specific to ubuntu platform. If you use debian, I suggest use the general one.

1 Like

Problem with the SSL CA cert

anyone on dev team is looking at that issue? Seafile Pro 6.2 is unusable with S3 backend

1 Like

What Linux distribution do you use? Have you tried both builds, one for CentOS and one for Ubuntu?

it’s Debian, and install is x86-64.tar.gz as always. If it’s generic linux it should not depend on thing like libnsspem which support is questionable and not available on other platforms. I will try ubuntu build later today.

ubuntu version works fine

Ubuntu version in debian 9.
Backend SSL issue gone , no furthur more issues been found until now.
Seaf-fuse stll left unstable.
So why not combine these two system server software into a common thing.
Ps , if you use google chrome,in the admin page,statics tab , you’ll see the file/space/users form are all jumping because of the css designed not in harmony with chrome , safari is ok.

Newest found,webdav is dead again…unavailable…

Newest found,again…

Can’t send users notifications/reset passwords,etc,which is controlled by backgroud-send-mail-task.Without any mail configuration issue in seahub_settings.py and seafevents.conf.

1 Like

I have this same issue, was there ever any resolution here?

Yes,he solved it by using the Ubuntu version (as can be seen above).

How are the actives users calculated?

My statistics show 0 active users sometimes (sometimes even multiple days in a row) although some users are active (They logged in the Web UI or synced some files with the client). The File statistics show the correct activity for these days.