SWA doesn't show "new" mails

[direwolf]

Hello Community,

yesterday i did an upgrade/migration from v10 (old machine, FC4) to v11.0.3.31 (new machine, FC5). Migrating mails and settings from one server to the other worked just fine, after about 2 hours i had a running system.

Now my users report some very strange behaviour of Scalix which i could verify:

They only see Mails from before the migration in SWA Inbox, in Outlook they see all, even new mails, that were recieved after the migration yesterday.

Why this incosistency between using SWA and Outlook? Any Idea how to solve this?

Thanks a lot,
Michael

[ScalixSupport]

Hi Michael!

Can you try clearing the IMAP cache for one of the user and see if this helps for him/her
follow the steps below to clear IMAP cache:
1. make sure the user is logged out, i.e.

Code: Select all

omstat -u all

should not show any more sessions for any user.

2. determine the user folder directory, i.e.

Code: Select all

omshowu -n <lastname> -f

Note: <lastname> should be the lastname of the user in question.

3. Go there (the so-called g-directory of the user)

Code: Select all

cd /var/opt/scalix/??/s/user/????????/

4. This should have a subdirectory called imap-cache. delete that:

Code: Select all

rm -r imap-cache

5. Try accessing the user's mailbox using IMAP again.

Now, restart scalix-tomcat service and try again.

Thanks,
Subir

[voice_of_buddha]

This works here. Thanks Subir!

[direwolf]

Hello Subir,

thank you, seems to work.

And now, i have to do that for all my users ... *sigh* ,)

Michael

[racmar]

Here is a little program i wrote to help easily delete all the imap-cache folders. This will create a file called sxclearimap.sh in the current directory.

Code: Select all

#! /usr/bin/python
#
# program: sxclearimap.py
#  author: Robert Chavers
#   email: rchavers att aetco dott us
#
# purpose: create an output file called sxclearimap.sh
#          when sxclearimap.sh is executed, it will
#          delete all files located in every scalix users'
#          imap-cache folders
# WARNING: use at your own risk!


import os, sys, re
instance_cfg = '/etc/opt/scalix/instance.cfg'


re_datadir = re.compile("OMDATADIR=(.*?)\r?\n", re.IGNORECASE)
re_userfolder = re.compile('User Folder : ~/*(.*?)\r?\n', re.IGNORECASE)

# get scalix datadir location
omdatadir = re_datadir.findall(open(instance_cfg, 'r').read())[0]

# get a list of all scalix users
user_list  = os.popen('omshowu -m all','r').readlines()

# loop through all users to find their imap-cache folders
out = ['#!/bin/sh']
for user_line in user_list:
    user_name = user_line.split('/')[0].strip()

    # get this user's info
    userinfo = os.popen('omshowu -n "%s" -f' %user_name,'r').read()

    # find this user's imap-cache folder
    userfolder = re_userfolder.findall(userinfo)[0]
    userfolder = '/'.join(userfolder.split('/')[:-1])
    imapfolder = os.path.join(omdatadir, userfolder, 'imap-cache')

    # append the rm command to the output file
    out.append('rm -f %s/*' %imapfolder)

# extra blank line at the end
out.append('')

# write the output file to the current folder
open('sxclearimap.sh', 'w').write('\n'.join(out))


CHECK THIS FILE'S CONTENT BEFORE EXECUTING IT!

Code: Select all

cat sxclearimap.sh

When you are SURE it is safe, execute the output file.

Code: Select all

sh sxclearimap.sh

[ScalixSupport]

Hi!

One such script is available at Scalix Wiki page as well :
http://www.scalix.com/wiki/index.php?ti ... minScripts

Thanks,
Subir

[Trevor Benson]

Is there any other issue that can cause something like this? I have followed the directions 3 times but it doesnt resolve my issue. Although maybe mine is not exactly the same.

I have lots of subfolders under the Inbox. All of these folders update properly and show new mail as it arrived. However, the Inbox itself does not. The left pane file list shows an increase in inbox messages that matches the new message count, but the right pane NEVER shows me new inbox messages. This is the same from FF on Linux and IE on windows.

I am running Scalix 11.0.3. When i do a

Code: Select all

service scalix-postgres restart

then the messages that are missing shop up (if i dont close SWA). But subsequent emails do not show up. I have to restart scalix-postgress to get inbox, but subfolders work fine still.

The word above is supposed to be SHOW, but it keeps showing SHOP instead... not sure why....

Thanks,
Trevor

[KevinAnderson]

I'd suggest upgrading to 11.0.4. This issue was resolved in 11.0.3, but now that it's out, why not go right to 11.0.4.

Kev.

[Trevor Benson]

My concern is that I am already AT 11.0.3 with this issue persisting. So if the fix was there I would assume going to 11.0.4 wont make much more difference, although I have it on the server to install today, so shortly ill know.

Trevor

[Trevor Benson]

I have the same issue as before, although at least now Firefox is giving an error every time I sign into the mailbox. I am about to repeat the steps suggested before, but last time it just transfered a few messages and required repeating the step over and over to get new mail.

Here is the error provided from firefox now that I am running 11.0.4



And clicking send and receive repeats this over and over again.

Trevor

[bazzi22]

Is there any other issue that can cause something like this? I have followed the directions 3 times but it doesnt resolve my issue. Although maybe mine is not exactly the same.

I have lots of subfolders under the Inbox. All of these folders update properly and show new mail as it arrived. However, the Inbox itself does not. The left pane file list shows an increase in inbox messages that matches the new message count, but the right pane NEVER shows me new inbox messages. This is the same from FF on Linux and IE on windows.

I am running Scalix 11.0.3. When i do a

Code: Select all

service scalix-postgres restart

then the messages that are missing shop up (if i dont close SWA). But subsequent emails do not show up. I have to restart scalix-postgress to get inbox, but subfolders work fine still.

The word above is supposed to be SHOW, but it keeps showing SHOP instead... not sure why....

Thanks,
Trevor

Hi There,

I have the same issue, only when I restart the scalix-postgres the new messages are showing up.

I am running:

Code: Select all

rpm -qa|grep scalix
scalix-tomcat-5.5.16-263
scalix-res-11.0.4.25-1
scalix-mobile-11.0.4.25-1
scalix-tomcat-connector-11.0.4.25-1.rhel4
scalix-server-11.0.4.10790-1.rhel4
scalix-platform-11.0.4.25-1
scalix-swa-11.0.4.25-1
scalix-sis-11.0.4.25-1
scalix-postgres-11.0.4.25-1
scalix-sac-11.0.4.25-1


What can I do about it?

[mshade]

I have a user experiencing this problem now - we also migrated from 10 - but to 11.0.4 instead of 11.0.3.

Clearing imap cache fixed the problem temporarily yesterday, but today it returned. Today, I took a look at the postresql pgstartup.log as well, and saw some entries like this:

Code: Select all

ERROR:  invalid byte sequence for encoding "UNICODE": 0x00
ERROR:  current transaction is aborted, commands ignored until end of transaction block
ERROR:  invalid byte sequence for encoding "UNICODE": 0x00
ERROR:  current transaction is aborted, commands ignored until end of transaction block
ERROR:  invalid byte sequence for encoding "UNICODE": 0x00
ERROR:  current transaction is aborted, commands ignored until end of transaction block
ERROR:  invalid byte sequence for encoding "UNICODE": 0x00
ERROR:  current transaction is aborted, commands ignored until end of transaction block


After logging in with SWA, I noticed that some of the messages had corrupt 'from' addresses, in Unicode.

It looks like this may be part of the issue. Any other thoughts?[/code]

[mshade]

We are now on 11.2 and the problem is persisting. It's truly a pain to be walking on eggshells, waiting for another call from a user on the road who now can't read their new mail.

Clearing the imap cache and restarting scalix-tomcat seems to work consistently, but why is it getting corrupt in the first place?

In the case of one affected user, they use only webmail exclusively.

How can we track down what's causing this?

[lleahu]

I have a customer that is also experiencing this problem. Any help on this would be appreciated.

[dboreham]

I have a user experiencing this problem now - we also migrated from 10 - but to 11.0.4 instead of 11.0.3.

Clearing imap cache fixed the problem temporarily yesterday, but today it returned. Today, I took a look at the postresql pgstartup.log as well, and saw some entries like this:

Code: Select all

ERROR:  invalid byte sequence for encoding "UNICODE": 0x00
ERROR:  current transaction is aborted, commands ignored until end of transaction block
ERROR:  invalid byte sequence for encoding "UNICODE": 0x00
ERROR:  current transaction is aborted, commands ignored until end of transaction block
ERROR:  invalid byte sequence for encoding "UNICODE": 0x00
ERROR:  current transaction is aborted, commands ignored until end of transaction block
ERROR:  invalid byte sequence for encoding "UNICODE": 0x00
ERROR:  current transaction is aborted, commands ignored until end of transaction block


After logging in with SWA, I noticed that some of the messages had corrupt 'from' addresses, in Unicode.

It looks like this may be part of the issue. Any other thoughts?[/code]

Hi, I believe you are seeing this bug :
http://bugzilla.scalix.com/show_bug.cgi?id=14282

There is information on how to resolve the issue here :
http://www.scalix.com/wiki/index.php?ti ... 03-PGCHARS