Daily and Disk backup - hang

[Fred Rat]

Running PIAF 2.0.6.2, purple 32b, FreePBX 2.9.0.9, asterisk 1.8.6.
ran the install-dailybackup and install-diskbackup, edited the two config files in /etc/PBX.

When I run either of the scripts in /etc/cron.daily or /etc/cron.weekly to test, the scripts seem to hang, PS listing below.

I've not found anything useful in /var/log and don't know where else to look.

Please advise on suggestions on either what I did wrong, or where to look for more info, or perhaps even that these scripts aren't working yet on 2.0.6.2.


root@pbx:~ $ ps auxwww | grep back
root 2708 0.0 0.0 1888 296 ? TN Jan30 0:00 /etc/cron.daily/daily-backup.cron
root 2709 0.0 0.0 2520 700 ? SN Jan30 0:00 awk -v progname /etc/cron.daily/daily-backup.cron progname {????? print progname ":\n"????? progname="";???? }???? { print; }
root 2710 0.0 0.0 0 0 ? ZN Jan30 0:00 [daily-backup.cr] <defunct>
root 5864 0.0 0.0 1868 300 ? TN Feb01 0:00 /etc/cron.weekly/disk-backup.cron
root 5865 0.0 0.0 2520 732 ? SN Feb01 0:00 awk -v progname /etc/cron.weekly/disk-backup.cron progname {????? print progname ":\n"????? progname="";???? }???? { print; }
root 5866 0.0 0.0 0 0 ? ZN Feb01 0:00 [disk-backup.cro] <defunct>
root 30457 0.0 0.0 4328 732 pts/1 S+ 08:53 0:00 grep back

 

[darmock]

there should be logs for both programs in /var/log if not then there are other problems. try running the programs in the cli with the -v command line and report the version back.

Also we NEVER supported asterisk 1.8.6 with PBX in a Flash 2.0.6.2.1 The earliest is 1.8.7.2 so what is the unedited output from status? Did you run the update-source program and downgrade to an earlier *unsupported* version
of asterisk.


Tom

 

[Fred Rat]

Sorry, I had the asterisk version wrong, it's 1.8.8.0

doing a ./daily-backup.cron -v
yields a stuck process
root 5026 0.0 0.0 1888 296 pts/0 T 14:58 0:00 ./daily-backup.
cron -v
root 5027 0.0 0.0 0 0 pts/0 Z 14:58 0:00 [daily-backup.c
r] <defunct>

 

[darmock]

My mistake daily-backup.cron -ver

I have the flu right now so I am only firing on 1/2 cylinder.


Are there no log files?


Tom

 

[Fred Rat]

No log files
root@pbx:/etc/cron.daily $ ./daily-backup.cron -ver

[2]+ Stopped ./daily-backup.cron -ver

root@pbx:/etc/cron.daily $ ls -l /var/log/*aily*
-rw-r--r-- 1 root root 14753 Jan 29 12:09 /var/log/install-dailybackup.log
root@pbx:/etc/cron.daily $

Sorry about the flu.. ChickenSoup to you.

??

 

[darmock]

you are not using the correct version of the cron files. Some you managed to get installed the 1757 version or one that is not specifically engineered for your version. Unfortunately centos 6 has some odd problems which is why we keep 2 separate versions one 32 and one 64. How did you install the programs?

Please post an unedited version of the output from status.

Tom

 

[Fred Rat]

Ran the install-dailybackup script See listing from "history" below. How do I fix?

1 nslookup yahoo.com
2 yum update
3 update-programs
4 update-fixes
5 ls
6 top
7 df
8
9 ping
0 status
1 yup update
2 yum update
3 top
4 install-dailybackup
5 cd /etc/
6 ls -ltr
7 cat /etc/crontab
8 cd pbx/
9 ls
0 cat daily-backup.conf

 

[darmock]

cant help without an unedited copy of the status program. Really this is not tough type status from the command line capture the screen and post if.

BTW You ran yum update which is never ever recommended unless you do it through update-source. However your box to break as you see fit.


Tom

 

[Fred Rat]

Did not realize that a raw yum was a bad idea, may be my box to break, but I'ld rather not, so thanks for that tip.

status output .txt has been attached, only minor edit to hide some data done to it.

Thanks much for your help btw.

 

[darmock]

1. update-programs -d

2. remove the existing /etc/cron.daily/daily-backup.cron file also remove the /etc/pbx/daily-backup.conf

3. pbx-menu

4. select install-dailybackup

5. edit the /etc/pbx/daily-backup.conf file

6. Try running ./etc/cron.daily/daily-backup.cron again


report back - follow steps exactly including -d after update-programs

Tom

Just tested on current 32/64 bit builds under centos 6.2 on systems installed with PIAF 20621 ISO and all works

 

[Fred Rat]

Did a reboot, then did all of the steps except for 5 (edit the config) figured I would leave it unconfigured and see what happened.
ran daily-backup.cron with and without -ver and both cases got a defunct proc, and no log in /var/log except for the install log.

 

[darmock]

Well there would be no log as the file can't run. Tested this extensively and cannot duplicate it on our reference platforms so cant help. Works in our usual test matrix 32/64 bit running red/purple on real/proxmox/vmwareworkstation 8 So something is up with your install. I have no idea Below is results from one of our test systems

It could be something happened during yum update beyond that I have no idea. All of our test machines are up to date in terms of yum.

good luck

Tom

 

[Fred Rat]

I was suspicious of the yum as well, so I had a spare server and started all over with a fresh install. Did the install, updated the freepbx modules, and then using pbx-menu did the install of daily-backup. Let that finish and did a daily-backup -ver and it worked. Great, on to the next step..
Again using pbx-menu installed the full-disk backup, let that complete, tested with a disk-backup.cron -ver and that gave me a defunt process in PS.
Went back and retested the daily-backup with a -ver and now it fails and give a defunc process.
Just to be double sure, I did a reboot between running the disk-backup and the daily-backup to make sure things were clean.
It appears that something the disk-backup is doing during it's install is hosing the dailybackup.
Hope that gives you an idea or two.
Again thanks for your help.

Anyway to uninstall and/or fix the disk-backup?
(I would rather not go through another full install to get back to that point, but will if I have to)
Also, if I'm not supposed to do a yum, how exactly am I supposed to keep the Linux modules on the system up to date?

 

[darmock]

1. Why in the world do you want to constantly update a system when there is no real need? All you do is introduce new variables which make it difficult to track down problems. The only time you should update a stable running system is when the has been a patch issued for a real and verified security problem. We generally will tell you in our forums.

2. You can update using yum within update-source. This is where you can update kernels also if there is a real need. Once again doing this may introduce real problems with your pbx. I know we have been considering going the way of some of the other distros and maintaining our own repo so this crap does not happen,

Beyond that will look at what is going on and see if there in an interaction between daily and whole disk.

no way to uninstall. have no idea where the problem is hopefully I can duplicate your problem here if not.......

Tom

 

[darmock]

try

install-diskbackup -d

then run

./disk-backup.cron -ver

bet it works.


That being said I cannot duplicate the problem with after you install disk backup the daily backup fails to run.

I suspect some of our download mirrors are out of sync with current releases and have initiated a resync. This is the only explanation I can come up with should be fixed by tomorrow in the meantime you can try using the -d flag for the install-dailybackup also and see if it corrects the problem


Tom

 

[Fred Rat]

Success, I had to do a install -d for both the disk backup and the daily, but as I re-did each of them, and tested, it fixed the problem.

Thanks much for the help, I know this took a lot of message exchange and time. I do appreciate it and the very fine tools!