hawq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ming Li <...@pivotal.io>
Subject Re: problem on AMD FX-8350
Date Thu, 07 Jul 2016 09:31:13 GMT
I google this error, it seems this problem exists for a long time for
hawq/greenplum with CentOS 7.2.

----------------------------------------------------------------

Later versions of systemd will clear out semaphores on user logout by
default, which can interfere with hawq.

Our solution to get hawq to install on CentOS 7.2 is to run the following
as root before running hawq init:

echo "RemoveIPC=no" >> /etc/systemd/logind.conf
service systemd-logind restart

On Fri, Jul 1, 2016 at 4:08 PM, Lei Chang <lei_chang@apache.org> wrote:

> can you run "hawq check" on your environment to see whether OS parameters
> are correctly set?
>
> Thanks
> Lei
>
>
>
> On Fri, Jul 1, 2016 at 2:05 PM, xuebao.lu@seaboxdata.com <
> xuebao.lu@seaboxdata.com> wrote:
>
> > Hello,guys. help!!!
> > My cluster includes 4 virtual machines(ambari, master, slave1, slave2)
> > which are equipped with : the operating system 'Centos7.2-1511', memory
> of
> > 4G, and processor 'AMD FX-8350' for each one.
> > I've installed HAWQ using Ambari as follows :
> > - ambari : ambari-server, mysql;
> > - master : HDFS namenode, HDFS datanode, hawq-master, hawq-segment;
> > - slave1 : HDFS datanode, hawq-standby, hawq-segment;
> > - slave2 : HDFS datanode, hawq-segment.
> > After installation, the service hawq-standby could not be started. The
> > error log is as follows :
> > 192.168.86.101 50612 2016-06-30 17:12:51 CST 0 seg-10000
> > FATAL XX000 semctl(37093455, 11, SETVAL, 0) failed: Invalid argument
> > (pg_sema.c:154)
> > 0 pg_sema.c
> > 154 "Stack trace:
> > 1    0x87448a postgres errstart + 0x22a
> > 2    0x74bb3c postgres <symbol not found> + 0x74bb3c
> > 3    0x7a129f postgres InitProcess + 0x33f
> > 4    0x7affdd postgres PostgresMain + 0x8ad
> > 5    0x763c43 postgres <symbol not found> + 0x763c43
> > 6    0x76439d postgres <symbol not found> + 0x76439d
> > 7    0x7661ce postgres PostmasterMain + 0xc7e
> > 8    0x6c02da postgres main + 0x48a
> > 9    0x7f7ed3881b15 libc.so.6 __libc_start_main + 0xf5
> > 10   0x4a1709 postgres <symbol not found> + 0x4a1709
> >
> > For more info : i've tried to do the same installation in another
> cluster.
> > All the equipement is the same, except that the processor is Inter
> i5-4590.
> > In this situation, all the service could be started without any problem
> > after installation.
> >
> > Do you have any idea ?
> > Thank you in advance.
> >
> >
> > xuebao.lu@seaboxdata.com
> >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message