Printable View
Have you actually tried any of the solutions mentioned in the forums and if so, which ones have you tried? How about some details of the Zimbra version you have installed? There's a lack of detail in your post and a 'me too' doesn't really tell us much.Quote:
Originally Posted by unionsport
Hi,
for evaluation purpose, I've installed the new ZCS 7.0 and ran into the same problems as described in this thread.
I've tried all the mentioned suggestions, but the Server Status page on the Administration Console keeps reporting "stopped" for all Services.
How does the Server Staus page actually gather the information? Does it only rely on entries in /var/log/zimbra.log? If so, which ones?
My setup:
All packages except the zimbra proxy and the zimbra memcached package are installed on one machine. Software Versions:
Release 7.0.0_GA_3077.RHEL5_64_20110127201852 CentOS5_64 FOSS edition.
Instead of syslog that installes as default with centos 5.5, I am using rsyslog.
Thanks in advance for additional hints to pinpoint the problem.
Cheers
Sam
Ok, as there are quite a lot of posts about the stats problem and none of them has had an answer that was useful to me, I traced it myself.
First zmlogger does not support the timestamp format defined in RFC 3339, which since version 3 of rsyslog is the default.
Because of this, zmlogger wasn't able to catch any log message at all and hence the sqlitedb that gets queried by GetServiceStatusRequest was completely empty.
To overcome this, you either have to run rsyslod in backwards-compatibility mode or use the RSYSLOG_TraditionalFileFormat template. Alertnatively you can also adjust the logregex pattern of zmlogger, but this probably will be unsupported:
Beside of this limitation there are at least two bugs that can prevent the proper update of the sqlite db if you use rsyslogd on CentOS 5 (or RHEL 5):Code:525 my $logregex = qr/(^.{32}) ((\d+\.\d+\.\d+\.\d+)|(\S+)) ([^[]+)(\[(\d+)\])?: (.*)$/o;
- zmsyslogsetup does not reload rsyslogd after adjusting the rsyslogd.conf file. quick and dirty fix is to adjust zmsyslogsetup accordingly:
Code:485 my $syslog_PID=`cat /var/run/*syslog*.pid`;
- logrotate does not reload rsyslog after the roration of the zimbra.log and zimbra-stats.log file. hence the new created files stay empty until next time rsyslog get reloaded. adjust the lines in the postrotate section of /etc/logrotate.d/zimbra accordingly:
Code:7 kill -HUP `cat /var/run/rsyslog*.pid 2> /dev/null` 2> /dev/null || true
19 kill -HUP `cat /var/run/rsyslog*.pid 2> /dev/null` 2> /dev/null || true
Cheers
Sam
Hi All, I have OS Centos 5.5 X64 and ZCS 7.0.0_GA_3077. The same bug, using syslog-ng after tunning '' zimbra added some lines to /etc/syslog-ng/syslog-ng.conf, after syslog-ng restarting files /var/log/zimbra-stats.log /var/log/zimbra.log updates well and /etc/init.d/zimbra status shoes that all services running well but in web interface in server stats I see only red boxes. ssh keys already regenerated.
no errors in zimbra log files.
Any ideas?
did you tried the fixes told here in this thread?Quote:
Originally Posted by greggy
You talk about #45(http://www.zimbra.com/forums/install...tml#post208944) ?
my syslog-ng.conf:#1573865 - Pastie
in #45 1st fix allow rsyslog handle messages correct but I'm using syslog-ng, 2nd and 3d fixes make auto restart of rsyslog but I reload it many times and daemon uses new config.
ssh keys already renewed manually.
Any other fixes?
check the output of
zmprov gcf zimbraLogHostname
I think it point to wrong server, where no logger service installed. To change that do as user zimbra
zmprov mcf zimbraLogHostname correct.server.name
could you run zmsyslogsetup?
About hostname
------------------
Before setup ZCS I configure my zone(somehostname.org) on bind DNS server, all installer check's passed well. Trough install time hostname didn't change, zimbra daemons(slapd, postfix, etc) uses somehostname.org to bind and it works well.
# su - zimbra -c "zmprov gcf zimbraLogHostname"
zimbraLogHostname: somehostname.org
# nslookup somehostname.org
Server: 10.0.7.44
Address: 10.0.7.44#53
Name: somehostname.org
Address: 10.0.7.55
# ping -c 1 somehostname.org
PING somehostname.org (10.0.7.55) 56(84) bytes of data.
64 bytes from somehostname.org (10.0.7.55): icmp_seq=1 ttl=64 time=0.017 ms
--- somehostname.org ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.017/0.017/0.017/0.000 ms
# ip a show eth0
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast qlen 100
link/ether 55:55:55:55:55:55 brd ff:ff:ff:ff:ff:ff
inet 10.0.7.55/8 brd 10.255.255.255 scope global eth0
inet6 fe80::6ab5:99ff:fec9:20be/64 scope link
valid_lft forever preferred_lft forever
# zmprov mcf zimbraLogHostname somehostname.org
Doesn't help.
But zimbra parse logfiles with perl:
# lsof /var/log/zimbra*
COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
perl 12341 zimbra 3r REG 104,5 15793676 130632 /var/log/zimbra-stats.log
perl 13026 zimbra 3r REG 104,5 58044 130670 /var/log/zimbra.log
# ps aux | grep -E "12341|13026" | grep -v grep
zimbra 12341 0.0 0.1 42856 10956 ? Ss 11:01 0:00 /opt/zimbra/libexec/logswatch --config-file=/opt/zimbra/conf/logswatchrc --use-cpan-file-tail --pid-file=/opt/zimbra/log/logswatch.pid --daemon --script-dir=/opt/zimbra/data/tmp -t /var/log/zimbra-stats.log
zimbra 13026 0.0 0.1 42856 11748 ? S 11:01 0:00 /usr/bin/perl /opt/zimbra/data/tmp/.swatch_script.12994
About /opt/zimbra/libexec/zmsyslogsetup
----------------------------------------------
Already used and that script changed my /etc/syslog-ng/syslog-ng.conf after what /var/log/zimbra-stats.log /var/log/zimbra.log files appeared and syslog-ng writes msg from zimbra to it.
Any ideas?
Hi all,
I am experiencing the same problem as most of the other people in this thread -- when I go to the Zimbra Web Administration Console all of the items on the "Server Status" page have a red X next to them, even though all Zimbra services are running.
I'm running Zimbra 7.0.1 GA open source edition on my CentOS 5.5 dedicated server (I ran --platform-override during installation as I had downloaded the RHEL5 installer). Everything else in Zimbra is running perfectly except this.
i've tried some suggestions found on other threads in this forum, including re-generating SSH keys, deplying the new keys, running zmsyslogsetup, etc. All of those things worked but there has been no change to the "Server Status" page. The "Sever Statistics" page reports an error about being unable to read data from logger.
I've heard about some invalid configuration entries in syslog-ng.conf being the cause of the problem, but I don't have syslog-ng anywhere on my system so can't verify that. I don't know what Zimbra uses for logging on my server, but it's certainly not syslog-ng (although I do have both rsyslog and syslog).
I would really appreciate it if anyone has found a fix for this, or if anyone can provide me with suggestions on how I might be able to fix the problem. It's really annoying seeing all the servers being down when you know they're running smoothly (a "zmcontrol status" as the zimbra user confirms this).
Thanks for any help provided!