Koozali.org: home of the SME Server
Contribs.org Forums => Koozali SME Server 10.x => Topic started by: zeeclor on April 17, 2021, 04:26:01 AM
-
I have done a fresh install of SME10rc1 and cannot access horde.
# cat /var/log/horde/horde.log
2021-04-16T15:45:02+00:00 ERR: HORDE The database configuration is missing. [pid 749 on line 161 of "/usr/share/pear/Horde/Core/Factory/Db.php"]
I can add this to the bug tracker (I think) if this is not a known issue. I cannot see it listed anywhere however.
-
Some more info will help, like was a backup restored, VM, barmetal..
Over a number of sme10 installs beta and RC have not seen that error, there are errors in the error.log howebver webmail is still working, Kronolith is not being nice..there is an update pending
-
just fresh install ? any restore ?
what gives
ll /etc/e-smith/sql/init/
and
systemctl status -l mysql.init.service
and
systemctl status -l mariadb.service
-
just fresh install ? any restore ?
Fresh install on KVM with iscsi thin LUNs. I restore via affa but this is a plain install straight out of the box.
what gives
ll /etc/e-smith/sql/init/
[root@sme10rc1bck ~]# ll /etc/e-smith/sql/init/
total 0
and
systemctl status -l mysql.init.service
[root@sme10rc1bck ~]# systemctl status -l mysql.init.service
mysql.init.service - Koozali SME Server mysql DB injector for mariadb
Loaded: loaded (/etc/rc.d/init.d/mysql.init; enabled; vendor preset: enabled)
Active: active (exited) since Sat 2021-04-17 05:00:43 AEST; 8min ago
Process: 2676 ExecStart=/etc/rc.d/init.d/mysql.init start (code=exited, status=0/SUCCESS)
Main PID: 2676 (code=exited, status=0/SUCCESS)
Memory: 0B
CGroup: /system.slice/mysql.init.service
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au mysql.init[2676]: Loading horde_mysql_create_tables into mysql [ OK ]
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au mysql.init[2676]: Loading turba_mysql_create_objects into mysql [ OK ]
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au mysql.init[2676]: Loading kronolith_db_upgrade into mysql [ OK ]
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au mysql.init[2676]: Loading kronolith_utc_upgrade into mysql [ OK ]
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au mysql.init[2676]: Loading Ingo_Datatree_to_SQL into mysql [ OK ]
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au mysql.init[2676]: Loading Kronolith_Datatree_to_SQL into mysql [ OK ]
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au mysql.init[2676]: Loading Mnemo_Datatree_to_SQL into mysql [ OK ]
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au mysql.init[2676]: Loading Nag_Datatree_to_SQL into mysql [ OK ]
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au mysql.init[2676]: Loading Turba_Datatree_to_SQL into mysql [ OK ]
Apr 17 05:00:43 sme10rc1bck.zeeclor.com.au systemd[1]: Started Koozali SME Server mysql DB injector for mariadb.
and
systemctl status -l mariadb.service
[root@sme10rc1bck ~]# systemctl status -l mariadb.service
● mariadb.service - MariaDB database server
Loaded: loaded (/usr/lib/systemd/system/mariadb.service; enabled; vendor preset: enabled)
Drop-In: /usr/lib/systemd/system/mariadb.service.d
└─50koozali.conf
Active: active (running) since Sat 2021-04-17 05:00:41 AEST; 8min ago
Process: 2986 ExecStartPost=/usr/libexec/mariadb-wait-ready $MAINPID (code=exited, status=0/SUCCESS)
Process: 2954 ExecStartPre=/sbin/e-smith/systemd/mariadb-initialize (code=exited, status=0/SUCCESS)
Process: 2952 ExecStartPre=/sbin/e-smith/expand-template /etc/my.cnf (code=exited, status=0/SUCCESS)
Process: 2949 ExecStartPre=/sbin/e-smith/expand-template /root/.my.cnf (code=exited, status=0/SUCCESS)
Process: 2946 ExecStartPre=/sbin/e-smith/expand-template /var/lib/mysql/set.password (code=exited, status=0/SUCCESS)
Process: 2939 ExecStartPre=/sbin/e-smith/service-status mariadb (code=exited, status=0/SUCCESS)
Main PID: 2985 (mysqld)
Memory: 106.9M
CGroup: /system.slice/mariadb.service
└─2985 /usr/libexec/mysqld --defaults-file=/etc/my.cnf --basedir=/usr --datadir=/var/lib/mysql --user=mysql --pid-file=/var/run/mariadb/mariadb.pid
Apr 17 05:00:34 sme10rc1bck.zeeclor.com.au systemd[1]: Starting MariaDB database server...
Apr 17 05:00:34 sme10rc1bck.zeeclor.com.au mariadb-initialize[2954]: 210417 5:00:34 [Note] /usr/libexec/mysqld (mysqld 5.5.68-MariaDB) starting as process 2956 ...
Apr 17 05:00:39 sme10rc1bck.zeeclor.com.au mysqld[2985]: 210417 5:00:39 [Note] /usr/libexec/mysqld (mysqld 5.5.68-MariaDB) starting as process 2985 ...
Apr 17 05:00:41 sme10rc1bck.zeeclor.com.au systemd[1]: Started MariaDB database server.
The main issue seems to be a Mysqli.php issue.
[root@sme10rc1bck horde]# cat error.log
Apr 17 04:49:09 sme10rc1bck HORDE: SQL QUERY FAILED: Table 'horde.horde_perms' doesn't exist SELECT COUNT(*) FROM horde_perms WHERE perm_name = 'horde' [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:09 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'imp_sentmail_seq' DROP TABLE `imp_sentmail_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:10 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'ingo_rules_seq' DROP TABLE `ingo_rules_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:10 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'ingo_shares_seq' DROP TABLE `ingo_shares_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:13 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'kronolith_shares_seq' DROP TABLE `kronolith_shares_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:13 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'kronolith_resources_seq' DROP TABLE `kronolith_resources_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:14 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Table 'horde.rampage_types' doesn't exist SELECT type_id, type_name FROM `rampage_types` WHERE type_name IN ('calendar','event') [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:14 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'turba_shares_seq' DROP TABLE `turba_shares_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:15 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Table 'horde.rampage_types' doesn't exist SELECT type_id, type_name FROM `rampage_types` WHERE type_name IN ('contact') [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:16 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'nag_shares_seq' DROP TABLE `nag_shares_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:16 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Table 'horde.rampage_types' doesn't exist SELECT type_id, type_name FROM `rampage_types` WHERE type_name IN ('task') [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:17 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'mnemo_shares_seq' DROP TABLE `mnemo_shares_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:17 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Table 'horde.rampage_types' doesn't exist SELECT type_id, type_name FROM `rampage_types` WHERE type_name IN ('note') [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:19 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'horde_groups_seq' DROP TABLE `horde_groups_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:19 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'horde_histories_seq' DROP TABLE `horde_histories_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:20 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'horde_perms_seq' DROP TABLE `horde_perms_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:20 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'horde_vfs_seq' DROP TABLE `horde_vfs_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
Apr 17 04:49:20 sme10rc1bck HORDE: [horde] SQL QUERY FAILED: Unknown table 'horde_muvfs_seq' DROP TABLE `horde_muvfs_seq` [pid 3380 on line 323 of "/usr/share/pear/Horde/Db/Adapter/Mysqli.php"]
and all the tables do exist in MariaDB
edit:added the quote tags at the right place for easier reading
-
the last log you pasted is not a bug this is just log noise of horde trying to migrate what was already migrated, by trying to drop tables that have been already dropped.
-
Fresh install on KVM with iscsi thin LUNs. I restore via affa but this is a plain install straight out of the box.
just need some clarifications here.
1- Have you already restored something with affa on it before the message appear?
2- Affa is not and has never been designed to restore from a different version.
This is not a supported nor encouraged way to do, you will break things.
Affa is for high availability and fast recovering from disaster from one server to another with the same version.
-
just need some clarifications here.
1- Have you already restored something with affa on it before the message appear?
No. This is the first boot after install.
2- Affa is not and has never been designed to restore from a different version.
This is not a supported nor encouraged way to do, you will break things.
Affa is for high availability and fast recovering from disaster from one server to another with the same version.
OK.
Thanks for the heads up on the horde database recreation, Jean-Phillippe. It sounds like horde itself is OK.
How can one fix the missing database configuration (as per horde.log above)?
David
I should add there is no entry point for horde.
"The requested URL /horde was not found on this server."
-
No. This is the first boot after install.
Then it would be simpler and faster to simpley do a clean install again..you are indicating that this is just that a clean install.
-
Then it would be simpler and faster to simpley do a clean install again..you are indicating that this is just that a clean install.
Yes. It's a clean install and reinstalling gives the same error.log but no horde.log at first boot. It is a consistent result across two installations with essentially the same configuration.
Bugs that are not reproducible are the most insidious. :-(
-
Your right there, yet to experience it from alpha to the RC..damn..this will take a bit of detective work..
So, I see from one of your posts "install on KVM with iscsi thin LUNs" can you give us a rundown on hardware and hypervisor...just plucking straws
-
Yes we'll need to do this in little steps so we can see exactly where this occurs.
Can you do a completely clean install and reboot prior to any attempt at restore and let us know what you see.
-
Correction: I misinformed the forum. It works out of the box. I apologise for the miscommunication at our end that led to this.
We will continue to investigate our possible affa --rise issue but I suspect it too may be a luser error.
-
Great to hear, affa may not be the best, how big is a standard backup, it is proving to be workable..
Added: one of the contribs that has been ported over and is looking very rpomising is https://wiki.koozali.org/Rsync-backup2disk#Configuration
Might be worth a look
-
We will continue to investigate our possible affa --rise issue but I suspect it too may be a luser error.
As per JPs comment above please be EXTREMELY careful with affa.
It absolutely is not designed to restore between versions and there is every chance things will break.
(There are quite a few things in affa in general which I have in hand to fix but just not had the time them finish them yet - systemd awareness being one :-( )
There are recognised methods for the upgrade and you really ought to try and stick to them to save hassle.
If you DO use affa make sure you specify that if you have any issues.
-
As per JPs comment above please be EXTREMELY careful with affa.
It absolutely is not designed to restore between versions and there is every chance things will break.
(There are quite a few things in affa in general which I have in hand to fix but just not had the time them finish them yet - systemd awareness being one :-( )
There are recognised methods for the upgrade and you really ought to try and stick to them to save hassle.
If you DO use affa make sure you specify that if you have any issues.
OK.
Having noted the above I can say an affa --rise worked well on Sunday for our 450 gig of data on 10rc1 for both the source and destination VMs.
I also note the warning that "Your server has too much data for a reliable backup to desktop" for the standard backup mechanism.
-
OK.
Having noted the above I can say an affa --rise worked well on Sunday for our 450 gig of data on 10rc1 for both the source and destination VMs.
I would only state "seemed" to work well, potentials issues are more subtle than a server broken immediately...
could be a rpm file with SME9 version instead of SME10, could be permissions on a file or folder ....
but as this is your choice to take this path, please in case of issue, report how you migrated your server.
I also note the warning that "Your server has too much data for a reliable backup to desktop" for the standard backup mechanism.
Well backup to desktop, in other words download a tar archive thru apache, is as a result limited in size, and is only one of the multiple mechanisms for backup.
The first you should use, is rather to connect to your server console, and use the backup console menu, to backup to USB, or attached disk
Another alternative could be the dar backup form the same menu you saw the message that the " backup to desktop" was not possible.
I will not unroll here the backup on tape ;)
All of this is explained in SME Server Manual.
https://wiki.koozali.org/SME_Server:Documentation:Administration_Manual:Chapter10
https://wiki.koozali.org/SME_Server:Documentation:Administration_Manual:Chapter6#Option_8:_Perform_backup_to_USB_device
-
I would only state "seemed" to work well, potentials issues are more subtle than a server broken immediately...
could be a rpm file with SME9 version instead of SME10, could be permissions on a file or folder ....
but as this is your choice to take this path, please in case of issue, report how you migrated your server.
Just the usual. affa --rise
I failed to appreciate that while data is data, Centos 7 is not Centos 6.
In terms of ongoing problems dhcpd won't start and I have cvm-unix.service failures. I wonder if the latter is due to gremlins in the system. I could not find much documentation on that.
Backup and restore through the web GUI using nfs to a workstation seems workable for our VMs. Emulating a USB drive looks to be a pain.
-
Just the usual. affa --rise
I failed to appreciate that while data is data, Centos 7 is not Centos 6.
indeed CentOS 7 is not CentOS 6
In terms of ongoing problems dhcpd won't start and I have cvm-unix.service failures. I wonder if the latter is due to gremlins in the system. I could not find much documentation on that.
dhcpd : update to last available e-smith-base...
cvm-unix : will need more to be able to help
Backup and restore through the web GUI using nfs to a workstation seems workable for our VMs. Emulating a USB drive looks to be a pain.
if you use a vm simply try to add another virtual disk as sata. most of the time the console backup sees it. We should change the USB term to removable...
-
dhcpd : update to last available e-smith-base...
Good to have dhcpd back.
cvm-unix : will need more to be able to help
The top entry is the crash report.
The second entry shows lots of cvm-unix logs for users unknown to our systems. I presume these are not external attempts to login but I cannot tell.
#journalctl -xe
-- Subject: Unit cvm-unix.service has finished start-up
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
--
-- Unit cvm-unix.service has finished starting up.
--
-- The start-up result is done.
Apr 27 21:17:15 mycompany.com cvm-unix[16708]: Starting.
Apr 27 21:17:21 mycompany.com kernel: cvm-unix[16708]: segfault at 0 ip 00007f8266b2bf27 sp 00007ffc76851c98 error 4 in libc-2.17.so[7f82669ed000+1c4000]
Apr 27 21:17:21 mycompany.com systemd[1]: cvm-unix.service: main process exited, code=killed, status=11/SEGV
Apr 27 21:17:21 mycompany.com systemd[1]: Unit cvm-unix.service entered failed state.
Apr 27 21:17:21 mycompany.com systemd[1]: cvm-unix.service failed.
Apr 27 21:17:22 mycompany.com systemd[1]: cvm-unix.service holdoff time over, scheduling restart.
Apr 27 21:17:22 mycompany.com systemd[1]: Stopped Credential Validation Modules.
-- Subject: Unit cvm-unix.service has finished shutting down
snip
-- The start-up result is done.
Apr 27 20:36:45 mycompany.com cvm-unix[15861]: Starting.
Apr 27 20:39:22 mycompany.com cvm-unix[15861]: - abinn@mycompany.com
Apr 27 20:39:27 mycompany.com cvm-unix[15861]: - abinn@localhost
Apr 27 20:40:01 mycompany.com crond[15883]: pam_unix(crond:session): session opened for user root by (uid=0)
Apr 27 20:40:01 mycompany.com CROND[15884]: (root) CMD (/usr/bin/php /usr/bin/horde-alarms)
Apr 27 20:40:02 mycompany.com CROND[15883]: pam_unix(crond:session): session closed for user root
Apr 27 20:40:22 mycompany.com cvm-unix[15861]: - anushia_caines@mycompany.com
Apr 27 20:40:28 mycompany.com cvm-unix[15861]: - anushia_caines@localhost
Apr 27 20:42:07 mycompany.com cvm-unix[15861]: - bright@mycompany.com
Apr 27 20:42:12 mycompany.com cvm-unix[15861]: - bright@localhost
Apr 27 20:43:06 mycompany.com cvm-unix[15861]: - victoria-carr@mycompany.com
Apr 27 20:43:11 mycompany.com cvm-unix[15861]: - victoria-carr@localhost
Apr 27 20:44:02 mycompany.com cvm-unix[15861]: - victoria-carr@mycompany.com
Apr 27 20:44:07 mycompany.com cvm-unix[15861]: - victoria-carr@localhost
Apr 27 20:44:16 mycompany.com cvm-unix[15861]: - anushia_caines@mycompany.com
Apr 27 20:44:23 mycompany.com cvm-unix[15861]: - anushia_caines@localhost
Apr 27 20:44:55 mycompany.com cvm-unix[15861]: - anushia_caines@mycompany.com
Apr 27 20:45:01 mycompany.com crond[15935]: pam_unix(crond:session): session opened for user root by (uid=0)
Apr 27 20:45:01 mycompany.com crond[15934]: pam_unix(crond:session): session opened for user root by (uid=0)
Apr 27 20:45:01 mycompany.com CROND[15936]: (root) CMD (/sbin/affa --silent --_cronupdate)
Apr 27 20:45:01 mycompany.com CROND[15937]: (root) CMD (/usr/bin/php /usr/bin/horde-alarms)
Apr 27 20:45:01 mycompany.com CROND[15935]: pam_unix(crond:session): session closed for user root
Apr 27 20:45:01 mycompany.com CROND[15934]: pam_unix(crond:session): session closed for user root
Apr 27 20:45:02 mycompany.com cvm-unix[15861]: - anushia_caines@localhost
if you use a vm simply try to add another virtual disk as sata. most of the time the console backup sees it. We should change the USB term to removable...
SME Server found no USB, SCSI, SATA or VirtIO discs in the console backup procedure for our iscsi vms on debian stable KMV machines
-
good news for dhcp
for cvm from what i read it works
it is only used by qpsmtpd and sqpsmtpd. so check if those attempts are correlated with log in qpsmtpd but i would say, yes they are
for virtual disk backup (sata mode) all i can say it works with proxmox 6 which is debian based and with virtio on fedora.
ah yes, seeing you did not update to last available for SME 10, for it to work you need the last version available of e-smith-backup.
if you did not update for a while, you will need to first update smeserver-yum with all repo disabled except smeupdates.
-
May be a non issue however, with my Prox VMs I have to have allocated a USB device in the hardware setup of the VM.