Hi Guo,
Sorry for my delayed response.
I see this in your Operation Logs:
"\"Mon Sep 10 14:57:48 +0800 2012
/Stage[1]/Hdp::Pre_install_pkgs/Hdp::Exec[yum install
$pre_installed_pkgs]/Exec[yum install $pre_installed_pkgs]/returns
(notice): ganglia-gmetad-3.2.0-99.x86_64 from HDP-UTILS-1.0.1.14 has
depsolving problems\"",
"\"Mon Sep 10 14:57:48 +0800 2012
/Stage[1]/Hdp::Pre_install_pkgs/Hdp::Exec[yum install
$pre_installed_pkgs]/Exec[yum install $pre_installed_pkgs]/returns
(notice): --> Missing Dependency: libganglia-3.2.0.0.so.0()(64bit) is
needed by package ganglia-gmetad-3.2.0-99.x86_64 (HDP-UTILS-1.0.1.14)\"",
"\"Mon Sep 10 14:57:48 +0800 2012
/Stage[1]/Hdp::Pre_install_pkgs/Hdp::Exec[yum install
$pre_installed_pkgs]/Exec[yum install $pre_installed_pkgs]/returns
(notice): *Error: Missing Dependency: libganglia-3.2.0.0.so.0()(64bit) is
needed by package ganglia-gmetad-3.2.0-99.x86_64 (HDP-UTILS-1.0.1.14)\"",*
"\"Mon Sep 10 14:57:48 +0800 2012
/Stage[1]/Hdp::Pre_install_pkgs/Hdp::Exec[yum install
$pre_installed_pkgs]/Exec[yum install $pre_installed_pkgs]/returns
(notice): You could try using --skip-broken to work around the problem\"",
"\"Mon Sep 10 14:57:48 +0800 2012
/Stage[1]/Hdp::Pre_install_pkgs/Hdp::Exec[yum install
$pre_installed_pkgs]/Exec[yum install $pre_installed_pkgs]/returns
(notice): You could try running: package-cleanup --problems\"",
"\"Mon Sep 10 14:57:48 +0800 2012
/Stage[1]/Hdp::Pre_install_pkgs/Hdp::Exec[yum install
$pre_installed_pkgs]/Exec[yum install $pre_installed_pkgs]/returns
(notice): package-cleanup --dupes\"",
"\"Mon Sep 10 14:57:48 +0800 2012
/Stage[1]/Hdp::Pre_install_pkgs/Hdp::Exec[yum install
$pre_installed_pkgs]/Exec[yum install $pre_installed_pkgs]/returns
(notice): rpm -Va --nofiles --nodigest\"",
This is strange because libganglia is part of HDP-UTILS-1.0.1.14 repo.
What do you see when you run "yum list | grep libganglia"? You should see:
libganglia.x86_64 3.2.0-99 HDP-UTILS-1.0.1.14
Also, there are known issues with the code on trunk that makes it unstable
(I just tried deploying a cluster with it, and as soon as the cluster came
up I started getting false alarms from the monitoring system due to
https://issues.apache.org/jira/browse/AMBARI-662).
Can you try the 0.9 branch? That is the most stable version we have:
http://incubator.apache.org/ambari/install-0.9.html
Yusaku
On Mon, Sep 10, 2012 at 2:20 AM, Guo Leitao <[email protected]> wrote:
> Hi Yusaku,
>
> I passed 3rd step (Assign Master Services to Hosts) according to your
> advice, but stop at step 7 in "Deployment Progress". The error message is
> "Cluaster install failed". Would you kindly help to check what's wrong with
> my configuration? Please find attatched hmc.log, /var/log/httpd/error_log
> and OpentionLogs output in the webpage.
>
> Thanks!
>
>
> Leitao
>
> 2012/9/10 Yusaku Sako <[email protected]>
>
>> Hi Guo,
>>
>> When you added hosts in Step 2, did you use FQDN of the hosts? The FQDNs
>> should match what's given by the hosts when you run "hostname -f".
>> I ask because of the previous issue that you encountered regarding
>> certificate generation, and this can lead to other problems.
>> If you did not use FQDN, then please try using the FQDN for the cluster
>> hosts. Also, please try accessing Ambari via the FQDN and not the IP
>> address (I say this because I see "referer" as an IP address, not FQDN).
>>
>> Regarding the error_log output, "Undefined index: KERBEROS_SERVER in
>> /usr/share/hmc/php/frontend/selectServices.php on line 188" should not be
>> happening so there's definitely something wrong.
>> If you could also post your /var/log/hmc/hmc.log file, that would help us
>> debug the issue for you.
>>
>> Thanks,
>> Yusaku
>>
>>
>>
>> On Sun, Sep 9, 2012 at 7:10 PM, Guo Leitao <[email protected]> wrote:
>>
>>> Hi Yusaku,
>>>
>>> I tried Firefox 15.0.1 and Chrome 21.0.1180.89 m, but all have failed.
>>>
>>> Here are the logs in "/var/log/httpd/error_log"
>>>
>>> [Mon Sep 10 10:07:49 2012] [error] [client 10.6.1.165] PHP Warning:
>>> rmdir(/var/run/hmc/clusters/myTestCluster/) [<a
>>> href='function.rmdir'>function.rmdir</a>]: No such file or directory in
>>> /usr/share/hmc/php/frontend/localDirs.php on line 64, referer:
>>> http://192.168.34.8/hmc/html/initializeCluster.php
>>> [Mon Sep 10 10:07:49 2012] [error] [client 10.6.1.165] ClusterName:
>>> myTestCluster, referer:
>>> http://192.168.34.8/hmc/html/initializeCluster.php
>>> [Mon Sep 10 10:08:33 2012] [error] [client 10.6.1.165] PHP Notice:
>>> Undefined index: KERBEROS_SERVER in
>>> /usr/share/hmc/php/frontend/selectServices.php on line 188, referer:
>>> http://192.168.34.8/hmc/html/initializeCluster.php
>>> [Mon Sep 10 10:08:33 2012] [error] [client 10.6.1.165] PHP Notice:
>>> Undefined index: ZOOKEEPER_SERVER in
>>> /usr/share/hmc/php/frontend/selectServices.php on line 188, referer:
>>> http://192.168.34.8/hmc/html/initializeCluster.php
>>> [Mon Sep 10 10:08:33 2012] [error] [client 10.6.1.165] PHP Notice:
>>> Undefined index: HBASE_MASTER in
>>> /usr/share/hmc/php/frontend/selectServices.php on line 188, referer:
>>> http://192.168.34.8/hmc/html/initializeCluster.php
>>> [Mon Sep 10 10:08:33 2012] [error] [client 10.6.1.165] PHP Notice:
>>> Undefined index: OOZIE_SERVER in
>>> /usr/share/hmc/php/frontend/selectServices.php on line 188, referer:
>>> http://192.168.34.8/hmc/html/initializeCluster.php
>>> [Mon Sep 10 10:08:33 2012] [error] [client 10.6.1.165] PHP Notice:
>>> Undefined index: HIVE_SERVER in
>>> /usr/share/hmc/php/frontend/selectServices.php on line 188, referer:
>>> http://192.168.34.8/hmc/html/initializeCluster.php
>>> [Mon Sep 10 10:08:33 2012] [error] [client 10.6.1.165] PHP Notice:
>>> Undefined index: TEMPLETON_SERVER in
>>> /usr/share/hmc/php/frontend/selectServices.php on line 188, referer:
>>> http://192.168.34.8/hmc/html/initializeCluster.php
>>>
>>> Thanks,
>>>
>>> Leitao
>>>
>>>
>>> 2012/9/10 Yusaku Sako <[email protected]>
>>>
>>>> Hi Guo,
>>>>
>>>> What browser are you using?
>>>> FYI, in general the developers have been testing Ambari using Firefox
>>>> 14/15 (but recent versions of Chrome should work - IE most likely will not
>>>> work).
>>>>
>>>> Thanks,
>>>> Yusaku
>>>>
>>>>
>>>> On Sun, Sep 9, 2012 at 10:43 AM, Guo Leitao <[email protected]>wrote:
>>>>
>>>>> Hi Hitesh,
>>>>>
>>>>> Thanks for your kindly help. It works! But I find a new question about
>>>>> ambari.
>>>>> When start ambari service successfully, I following the instruction of
>>>>> website "http://AMBARIMASTER/hmc/html/"; to deploy Hadoop.
>>>>> In the 3rd step, that is "Assign Master Services to Hosts", there is
>>>>> no any response of the button "Next". Seems there is a javascript error.
>>>>>
>>>>> - Leitao
>>>>>
>>>>>
>>>>> 2012/9/6 Hitesh Shah <[email protected]>
>>>>>
>>>>>> The error is likely due to a missing ssl cert.
>>>>>>
>>>>>> Try this in the following order:
>>>>>>
>>>>>> $service ambari stop
>>>>>> $sudo puppet master --no-daemonize --debug
>>>>>> [ this will run in the foreground so once it comes up , go ahead and
>>>>>> kill it once the logs stop scrolling ]
>>>>>> $service ambari start
>>>>>>
>>>>>> The intermediate step is to create the ssl certificate if it did not
>>>>>> get created or was deleted by mistake.
>>>>>>
>>>>>> -- Hitesh
>>>>>>
>>>>>> On Sep 6, 2012, at 1:47 AM, Guo Leitao wrote:
>>>>>>
>>>>>> > I'm new to Ambari, and this is my first to config Ambari.
>>>>>> >
>>>>>> > After installed ambari-agent, ambari andhdp_mon_dashboard according
>>>>>> to http://incubator.apache.org/ambari/install.html, I can not start
>>>>>> ambari service. Here is error output. Any more configuration to puppet
>>>>>> should I do? Thanks!
>>>>>> >
>>>>>> > # service ambari start
>>>>>> > Starting Ambari Installer
>>>>>> > Starting httpd: Syntax error on line 35 of /etc/httpd/conf.d/
>>>>>> > puppetmaster.conf:
>>>>>> > SSLCertificateFile: file
>>>>>> '/var/lib/puppet/ssl/certs/compute-34-09.local.pem' does not exist or is
>>>>>> empty
>>>>>> > [FAILED]
>>>>>> > Failed to start Ambari
>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
>
