Ensim released an update to Webppliance as of 12/31/2002 - Version 3.1.3-4

This update can be downloaded from ftp.ensim.com

We have tested the update against our own RPM's and they do not conflict. These include the squirrelmail centralization, phpMyAdmin update, and mysql upgrade.

Some pointers for those upgrading:

1) Reboot your server before installing. If your uptime is less than 128 days its not necessary, but its always good to start with a fresh system. The amount of frustration from a incomplete upgrade is much more than the loss of a uptime statistic.

2) Make sure you have chmod 755 wpinstall-u.sh

3) For those upgrading from 3.0 you will need to edit

/etc/virtualhosting/tmpdirs and change

/home/virtual:/home/virtual/FILESYSTEMTEMPLATE/.tmp

To

/home:/home/virtual/FILESYSTEMTEMPLATE/.tmp

(credit to Mouse / ZB-Dev for pointing this out)

4) Make SURE your postgresql is running correctly; make SURE service postgresql restart works properly.
(Doing a upgrade when postgresql isn't running correctly can produce a very colourful experience that can require quite a bit of repair. In other words, fix your postgresql if you're getting errors in /var/log/messages relating to it not running.)

5) Make sure you have mysqld shut down before installing *VERY IMPORTANT*. Not having mysqld shut down will bomb the installer, leaving possibly hanging mysqld processes.

NOTE: The installer seems to have regular problems with mysql / mysqld. If you end up with a mysql that won't restart, try:

ps -axuf | grep mysql

Then kill the process id of the parent mysql process. Mysqld should restart after that. This has happened on both 3.23.41 mysqld and my RPM's.


6) If you are running non-RH compatible mysql packages (such as ones downloaded from MySQL.com), uninstall them before running the update.

7) If your install fails, check the installer.err file in the installation directory. Make the necessary changes (i.e. uninstallation of conflicting RPM's), then restart the update via

./oem_installer.sh -c installer.cfg

The installer is still affected by this following issue:

12/31/02 23:16:25 : rpm: no packages given for install

If you run into this while trying to complete a failed install, rpm -e any package (i.e. phpMyAdmin) and rerun oem_installer.sh. The installer will progress thru in that case and finish the upgrade.

I will be adding more information as it occurs to me.

thanks for the tips

i am going to attempt this 2morrow... gotta RTFM too

Well the upgrade pretty much went smooth as silk ....

I killed mysql to get ready for the upgrade but everytime I would get ready to install I'd notice mysql was running again ... I was like WTF ... I didn't need this messing up things ... Then I realised my SIM monitoring was restarting it every time it saw it down ... Removed SIM from crond and I was good to start the install ...

When I ran the install script it aborted and the installer.err log said I had to download and install:
rpm-4.0.4-7x.18.i386.rpm (provided by RedHat 7.3)

Got the rpm installed and the rest went without a hitch ...

Copied the http_app.conf file with the new one and updated all the info ... restarted httpd and all was good and no errors ... Everything seems to be running great ...

I'm so happy ... No I can drink those beers I've been saving ... Didn't want to try this drunk

BTW ... Thanks GPAN ... all the rpm's you've done that I installed (mysql, php, mtrg, proftpd and any others I forgot) had no conflict with this upgrade for me

I had a conflict, it wouldnt install cause it was trying to reinstall php-4.2.2-ensim6 but conflicted with gpans 4.2.3. I just removed gpans and reinstalled them after update.

On another note, If you changed your webppliance user per my howto then the 'webppliance' service will fail to start after this update. No biggie though, you will need to redo the changes in
/usr/lib/opcenter/fastcgi/httpd-tmpl.conf

per the howto or just do

rm -rf /var/tmp/dynamic

and webppliance will come back up fine as user/group nobody

Other than that the update went smooth

Thought I might as well leave some feedback as well

Update on 2 plain ensim boxes with 3.1.1 went ok other then the MYSQL restart error which was easily fixed by

/etc/rc.d/init.d/mysqld restart
/etc/rc.d/init.d/webppliance restart


The one machine with the global squirrelmail by gpan had a problem, sqmail stopped working - this was fixed by simply reinstalling the sqmail rpm again.


Appliance Administrator 3.1.3-4 (admin)

All good !


Nice timing by ensim, just before the NY Parties began - seems like a few folks at ensim worked thru the holidays.

I never got this error, and i had gpan's rpm installed for php 4.3.0

i did have a few probs with mysql at times tho :/

php-4.2.2-2ensim6 apache-1.3.27-ensim1
I need to uninstall these rpms for the installation to work how do iu do this i gtet the error

php is needed by webppliance-mod_php-3.1.1-24
php is needed by phpMyAdmin-2.2.0-ensim4
php = 4.2.2-2ensim6 is needed by php-ftp-4.2.2-2ensim6
php = 4.2.2-2ensim6 is needed by php-manual-4.2.2-2ensim6
php = 4.2.2-2ensim6 is needed by php-mysql-4.2.2-2ensim6
php = 4.2.2-2ensim6 is needed by php-snmp-4.2.2-2ensim6
php = 4.2.2-2ensim6 is needed by php-sysvsem-4.2.2-2ensim6
php = 4.2.2-2ensim6 is needed by php-sysvshm-4.2.2-2ensim6
php = 4.2.2-2ensim6 is needed by php-gettext-4.2.2-2ensim6
/usr/bin/php is needed by php-devel-4.2.2-2ensim6

and

apache is needed by frontpage-5.0-ensim15
apache is needed by phpMyAdmin-2.2.0-ensim4
apache is needed by tomcat4-4.0.3-ensim1
apache is needed by mod_jk-1.3-1.0-1.4.0.2ensim2
apache = 1.3.27-ensim1 is needed by mod_ssl-2.8.12-ensim1
apache >= 1.3.1 is needed by apache-mod_fastcgi-2.2.10-1ensim11


what commands do i use etc

Cdoverlaw - rpm -e for uninstalling rpm's though I would backup everything before removing them.




On another note, the update seemed to have changed the .dtml files (modified siteadmin pages etc) so all those with customised dtml files better back them up as mine seem to have been fubar'd .

Oh and seems like php is not configured with mysql properly, as its not listed under the configuration given out by phpinfo() - how the hell could they forget mysql ?! Since my sites are giving me Call to undefined function errors.

Someone please tell me their php/mysql sites are working properly after the upgrade and that im just a idiot who messed up something

I got this problem:

01/01/03 10:03:32 : error: failed dependencies:
libgpm.so.1 is needed by links-0.96-2

Anyone know what it is?

my rpm seems to have so much dependencies any easy guide for me to install the new rpm?

[root@box1 extrarpms]# rpm -Uvh rpm-4.0.4-7x.18.i386.rpm
error: failed dependencies:
popt = 1.6.4 is needed by rpm-4.0.4-7x.18
rpm = 4.0.3 is needed by rpm-python-4.0.3-1.03
rpm = 4.0.3 is needed by rpm-build-4.0.3-1.03
rpm = 4.0.3 is needed by rpm-devel-4.0.3-1.03
librpm-4.0.3.so is needed by gnorpm-0.96-11
librpm-4.0.3.so is needed by rpm-python-4.0.3-1.03
librpm-4.0.3.so is needed by rpm-build-4.0.3-1.03
librpmbuild-4.0.3.so is needed by rpm-build-4.0.3-1.03
librpmdb-4.0.3.so is needed by gnorpm-0.96-11
librpmdb-4.0.3.so is needed by rpm-python-4.0.3-1.03
librpmdb-4.0.3.so is needed by rpm-build-4.0.3-1.03
librpmio-4.0.3.so is needed by gnorpm-0.96-11
librpmio-4.0.3.so is needed by rpm-python-4.0.3-1.03
librpmio-4.0.3.so is needed by rpm-build-4.0.3-1.03

You would update all of them recursively, i.e.

1) Download all the dependency RPM's that aren't in the /tmp/extrarpm's folder

2) rpm -Uvh rpm* gnorpm* .. (et al)

[root@box1 rpminstall]# rpm -Uvh rpm* popt*
error: failed dependencies:
librpm-4.0.3.so is needed by gnorpm-0.96-11
librpmdb-4.0.3.so is needed by gnorpm-0.96-11
librpmio-4.0.3.so is needed by gnorpm-0.96-11


ok where to find those missing ones.. i can't seems to find them..can help me a bit here

rpmfind.net ?

have you tried a --force or --nodeps ?

You can use www.rpmfind.net to find newer version of gnorpm. You could also find newer copy on www.rpm.org

I got this in installer.log after i try installing wpinstall-u.sh


Failed to install service RPMs.


In installer.err I got this

Installer version 1.0-17.
Installer error log:
Thu Jan 2 02:19:57 SGT 2003
01/02/03 02:21:14 : error: failed dependencies:
openssh = 2.9p2-7 is needed by openssh-askpass-gnome-2.9p2-7


I am sure i have installed openssh 2.9p2-7

Try

rpm -qa | grep openssh

and see what versions of the rpm's you do have installed.

[root@box1 extrarpms]# rpm -qa | grep openssh
openssh-clients-2.9p2-7
openssh-askpass-3.1p1-6
openssh-2.9p2-7
openssh-askpass-gnome-2.9p2-7
openssh-server-2.9p2-7


so what next?

i try installed wpinstall.sh again and i still get some error

Failed to install service RPMs.
Installation aborted.
Please review /root/installer.err for more information.


in installer.err

WEBppliance 3.1.3-4.
Installer version 1.0-17.
Installer error log:
Thu Jan 2 02:35:38 SGT 2003
01/02/03 02:37:11 : error: failed dependencies:
openssh = 2.9p2-7 is needed by openssh-askpass-gnome-2.9p2-7

All i had to do was upgrade to redhat 7.3 rpm and it works like a charm!

i was running 3.1.3 already, so the update was rather quick & painless! (for a change)

I also had gpans mysql and other rpm's installed. NO problems with those either.

how do i upgrade to 7.3rpm ?

OpenSSH version should be 3.1 anyways; here's the official RH rpm's - download the ones in the 7.2 section; install them together

https://rhn.redhat.com/errata/RHSA-2002-127.html

Actually, 3.1 has security vulnerabilities, and it should be upgraded to 3.5

RH backported the patches (you can explode the srpm)

whenever I download the installer in ascii or binary mode I get this error when I run it:

Unpacking the setup program. This might take few minutes...

gunzip: stdin: invalid compressed data--format violated
Unable to run oem_installer.sh

I'm using gunzip 1.3 (default one).

What to do next? :-)

Try redownloading - you got a corrupt installer.

I have download the rpms that u asked me to download from the redhat site..

when i tried install the 5 openssh files i get this

[root@box1 openshh]# rpm -Uvh *.rpm
Preparing... ########################################### [100%]
package openssh-askpass-3.1p1-6 is already installed


when i do this, i still get the same result

[root@box1 openshh]# rpm -qa | grep openssh
openssh-clients-2.9p2-7
openssh-askpass-3.1p1-6
openssh-2.9p2-7
openssh-askpass-gnome-2.9p2-7
openssh-server-2.9p2-7

this is getting headache

Upgrade the other 4

##################################################

WEBppliance 3.1.3-4.
Installer version 1.0-17.
Installer error log:
Thu Jan 2 09:57:19 GMT 2003
01/02/03 09:58:21 : error: unrecognized db option: "db3" ignored.
error: cannot open Packages index using db1 - No such file or directory (2)
01/02/03 09:58:21 : error: unrecognized db option: "db3" ignored.
..........

01/02/03 09:58:22 : The following rpms are required: bash SysVinit iputils xinetd ncftp perl popt tcp_wrappers telnet-server textutils fileutils mount gzip gawk mktemp filesystem util-linux glibc tar passwd grep sed cpio findutils initscripts glibc tar passwd grep sed cpio findutils initscripts sh-utils
01/02/03 09:58:22 : Required RPMs were not found to be installed on this computer. If this message is in error then please co$
01/02/03 10:00:39 : error: unrecognized db option: "db3" ignored.
error: cannot open Packages index using db1 - No such file or directory (2)
error: cannot open Packages database in /var/lib/rpm


---

there is about 100 lines of:
error: cannot open Packages index using db1 - No such file or directory (2)

---

Anyone have any ideas ?

I had heaps of problems upgrading, failed to install service rpm's in particular webppliance-mysql even though sim switched off and mysqld stopped took me several hours to work out all the bugs,main prob was that it updated the version before it ran into the mysql prob so the install script wouldn't let me run it again.
cd downloads and rpm -Uvh --replacepkgs *.rpm after that wait, I had the webppliance httpd.conf user prob ta for the notice now rfxn I just edited the user again to run it as ocwhttpd but it seems to be all up and running now only other prob I had was after the install I installed php4.3.0 which system did not seem to like at all so a downgrade to my old patched 4.2.3..
ohh it also corrupted my quota files ,another thing to watch had to create them again....
I will put in a few bug reports to ensim another month or so the upgrade will blow out 20 or 30 MB's like the 3.1.1 upgrade

Hey i thought that was just me, all my quota files got corrupt too. Had to 'rm' them, regenerate new quota db's with 'quotacheck' and resave all my sites.

gpan... you're awesome... your pointers are great -- upgrade was flawless. :-)

Can you explain in detail how you did that? It would appear that all of my quota files are screwed up as well. I don't have any idea where they're located at and how to regenerate them, so any pointers would be appreciated.

Thanks.

Currently running Ensim 3-1-1

login SSH as root
#cd /
#cd root
#mkdir ensim-3-1-3
#cd ensim-3-1-3

#ftp ftp.ensim.com
#login : anonymous
#password : 123@123.com

#cd /download/webppliance/linux/3.1/
#binary
(change download mode to binary)
#prompt
(disable prompting)
#mget *
(download all files to local drive)
#quit

#chmod +x *.sh

#sim –j
(Kill SIM monitor cron job)

#/etc/rc.d/init.d/mysqld stop
(Stop MySQL before running installation / upgrade script)

#./wpinstall-u.sh

(Running the script, there are a few prompts)
(Welcome screen – Type “y” to proceed)
(License screen – Type “y” to proceed)
(Type “y” to accept defaults)

Installation / Upgrade starts …
Errr ... Installation / Upgrade crashes out with the following error
################################################################################
#################
WEBppliance 3.1.3-4.
Installer version 1.0-17.
Installer error log:
Fri Jan 3 16:51:31 SGT 2003
01/03/03 17:01:40 :
Unfortunately, there are bugs in the versions of rpm shipped with RedHat 7.1 and 7.2, and their respective updates, that can cause the upgrade to fail to replace files that it should. In order to prevent this, this install script requires that you have at least rpm-4.0.4-7x.18.i386.rpm (provided by RedHat 7.3) installed or newer. Please upgrade to this version of rpm or newer, and then run the upgrade script again. To ease your upgrade, you can find the relevant RPMs in the tmp/extrarpms directory. Note that you may also need to upgrade glibc, popt, and any of the rpm-* RPMs (e.g. rpm-devel, rpm-python, etc.) that you already have installed along with rpm; you may ignore the other RPMs present in that directory.
################################################################################
#################

#cd /
#cd tmp
#cd extrarpms
#rpm –Uvh rpm-4.0.4-7x.18.i386.rpm

after upgrading of rpm-4.0.4-7x.18.i386.rpm, successfully, we go back and run the installation / upgrade again

#cd /
#cd root
#cd ensim-3-1-3
#./wpinstall-u.sh

Installation / Upgrade starts …
Preparing packages for installation...

After a few minutes i.e 25 minutes, the following is seen….

Congratulations. WEBppliance was successfully installed.
Downloaded files in download may now be deleted.
Goto http://xxxxxxxxxxx.com to setup your WEBppliance.

Logged in as admin and found that the version number has been upgraded to Ensim-3-1-3

But, we still have some outstanding issues ….

#cd /
#cd root
#cd ensim-3-1-3
#rm –fr ./download

Remove all files from the /root/ensim-3-1-3/download folder
These files are not needed anymore

#cd /
#cd /etc/httpd/conf/

#mv httpd_app.conf httpd_app.conf.old
Keeping a copy of the original httpd_app.conf file

#cp /usr/lib/opcenter/apache/install/httpd_app.conf ./

#vi httpd_app.conf

Replace all instances of LOCALIP and LOCALHOST with the Webppliance Server IP and Hostname respectively.

Edit httpd_app.conf to include all the customizations made

Do a comparison between httpd_app.conf and httpd_app.conf.old

All amendments to /etc/httpd/conf/httpd.conf saved and quit

#/etc/rc.d/init.d/httpd restart

Everything seems okay ..... .... I hope. ...

#sim –j
Add back SIM monitor cron job

#/etc/rc.d/init.d/mysqld start
Start back MySQL services

Tested login with central squirrelmail … tested ok.

Tested site http://xxxxxxxxxx.com …. tested ok.

Tested site with MySQL DB http://xxxxxxxxxxxxx.com … tested ok

Upgrade to 3.1.3 seems to be successful .... I will continue to monitor ....

Thank GOD + Heavens + Ensim + Gpan + Foggy + All other posts that I have READ ...

PS - Now my fingers are stiff from all that crossing ...

Hope this helps ....

I have tried upgrad (I guess) from 3.0.3-2.
The installation completed without giving an error and the end of the log file says :

01/04/03 15:14:34 : Congratulations. WEBppliance was successfully installed.
01/04/03 15:14:34 : Downloaded files in download may now be deleted.
01/04/03 15:14:34 : Goto http://xxxx.com to setup your WEBppliance.

when I go to my admin site I see that it still shows the 3.0.3 version.

on the ther hand, I was expecting the update to last about 1-2 hours as I have 100 domains. But the update completed instantly.

any ideas?

Did you restart webppliance after the install?

yes I did. But no effect.
After I received the "Preparing packages for installation... " notification, it only takes a few seconds to complete the installation.
I have 100 domains in the server so I guess it should take about an hour to complete it, right?

I just got a box two weeks ago. Got this problem when upgrading:

Performing pre-install check... (this may take a few minute[FAILED]

Failed to install service RPMs.

Installation aborted.
Please review /root/ensim-3-1-3/installer.err for more information.


I checked installer.err. The only errors I found were:


Starting httpd: Syntax error on line 345 of /etc/httpd/conf/httpd_app.conf:
Cannot add module via name 'mod_fastcgi.c': not in list of loaded modules
[FAILED]
error: execution of %post scriptlet from apache-mod_fastcgi-2.2.10-1ensim11 failed, exit status 1


and


[ SUCCESS ] Service database entry for mysql installed..
1 out of 1 hunk FAILED -- saving rejects to file /usr/bin/safe_mysqld.rej
Starting MySQL: [ OK ]


I have no clue what fastcgi even is. I went to the website, but it looks like I have to mess with Apache to upgrade fastcgi. I'm new - can anybody help?

Oh, and I'm upgrading from 3.1.1-31

Its commonly run into.

I can look at it if you want, please private message me.

I commented out anything that referenced to fastcgi in my httpd.conf file. This fixed the problem. I successfully upgraded to 3.1.3-41. Everything seems to be running normally. Is there any side effect to this? Why did I have fastcgi installed to begin with? (Sorry, this may be off topic)

I had fastcgi added to my http.conf after I did the latest Ensim upgrade. Though it did not cause any issues I removed it from there and so far so good. :-)

I also noticed that the latest httpd_app.conf no longer contains the reference to fastcgi. It's probably best to take the new httpd_app.conf file and apply all your custom changes again.

About my Compaq DL320 upgrade: It went perfectly well except for one thing: I block all ftp access in my firewall which called the upgrade script to stall because it downloads files from the ensim server using ftp. Disabling the firewall and restarting solved the problem.

I also excluded openssh from the upgrade as I already got a higher version running.

After the upgrade was finished successfully I had to restart httpd to make the server responsive again, but for the rest it went as smoothly. I keep my fingers crossed that it stays that way.

Oh one minor thing: The help screen did not work in squirrelmail. I had to upgrade the config.php files to fix it again.

Jan Derk

I get this message in installer-err!
-------------------------------
WEBppliance 3.1.3-4.
Installer version 1.0-17.
Installer error log:
Sat Jan 4 23:46:44 PST 2003
01/04/03 23:47:39 : rpm: no packages given for install
_______________



What does it mean?


R

Furthermore: it says 3.1.3-4 on Appliance!

I just found out that the installer.err log says: It seems that the new Ensim does install fastcgi and it does add the module to the current httpd_app.conf. However the fastcgi module is not activated in the Ensim 3.1.3-4 httpd_app.conf. I am confused :confused:.

So if you use a lot of cgi and performance goes down after this upgrade you might want to try to add the fastcgi module again to apache. I leave it out as I don't use cgi on my box.

Ok I attempted this.. and got this:



Well, I did EditVirtDomain domain.com in ssh, got no-where really:

If it has been marked for deletion, why hasn't it been deleted?

Can anyone suggest anything, how to remove it, or something?

Thanks in advance..,

WEBppliance 3.1.3-4.
Installer version 1.0-17.
Installer error log:
Mon Jan 6 17:08:34 CET 2003
01/06/03 17:11:43 : error: failed dependencies:
MySQL conflicts with mysql-3.23.41-1
MySQL-devel conflicts with mysql-devel-3.23.41-1


what can i do ? uninstall mysql ?

If you upgraded MYSQL yourself you have to exclude mysql from the upgrade. It's all in the fine manual that you read before you started the ensim upgrade. Right?

http://www.ensim.com/support/wpls/document..._ls_upgrade.pdf

Jan Derk

See the "How to manually delete a site" howto

thanks guys.