Page 2 of 4
Re: Vertica Nodes Randomly Fail
Posted: Wed Jul 17, 2013 3:23 am
by becky
Hi,
Unfortunately my node 3 failed again.
Here is my /opt/vertica/config/vspread.conf file:
Code: Select all
Spread_Segment XXX.XXX.XXX.255:4803 {
NXXXXXXXXX131 XXX.XXX.XXX.131 {
XXX.XXX.XXX.131
127.0.0.1
}
NXXXXXXXXX179 XXX.XXX.XXX.179 {
XXX.XXX.XXX.179
127.0.0.1
}
NXXXXXXXXX180 XXX.XXX.XXX.180 {
XXX.XXX.XXX.180
127.0.0.1
}
}
EventLogFile = /dev/null
EventTimeStamp = "[%a %d %b %Y %H:%M:%S]"
DaemonUser = spread
DaemonGroup = verticadba
Does that look right?
Re: Vertica Nodes Randomly Fail
Posted: Wed Jul 17, 2013 3:59 am
by becky
Hi Scutter,
Sorry, I messed up on the first attempt at reconfiguring the spread.
I re-ran the vertical_install like this:
- /opt/vertica/sbin/install_vertica -T -S default -s vertica01,vertica02,vertica03 -r vertica-6.1.2-0.x86_64.RHEL5.rpm
Although, after running the above command, the /opt/vertica/config/vspread.conf file did not change.
I'll let you know if the nodes stay up...
Re: Vertica Nodes Randomly Fail
Posted: Wed Jul 17, 2013 12:20 pm
by becky
Hi,
Bad news... woke up this morning and all three nodes were down
Seems like one failed, then later another failed bringing down the whole database.
This is baffling!
Re: Vertica Nodes Randomly Fail
Posted: Wed Jul 17, 2013 2:31 pm
by scutter
Try renaming the vspread.conf on all of the nodes and then re-running install_vertica again with -T -S default then check to see if you get the multiple Spread_Segments.
Re: Vertica Nodes Randomly Fail
Posted: Wed Jul 17, 2013 4:02 pm
by becky
Hi Scutter,
I renamed the /opt/vertica/config/vspread.conf to vspread.conf_old. Then I reinstalled Vertica. After the install, I checked the new vspread.conf file and it looks exactly the same as the old one.
Bummer...
Re: Vertica Nodes Randomly Fail
Posted: Wed Jul 17, 2013 4:21 pm
by scutter
When I have a chance, I'll try the reconfiguration on a couple of my VMs to see what I get.
Re: Vertica Nodes Randomly Fail
Posted: Thu Jul 18, 2013 1:24 am
by scutter
Hi Becky,
I created a 2-node cluster using 6.1.2 and with -T -S default with install_vertica, and I get the expected multiple Spread_Segments:
Spread_Segment 192.168.1.10:4803 {
N192168001010 192.168.1.10 {
192.168.1.10
127.0.0.1
}
}
Spread_Segment 192.168.1.12:4803 {
N192168001012 192.168.1.12 {
192.168.1.12
127.0.0.1
}
}
You renamed the vspread.conf on all three nodes before reinstalling? I suggest continuing to look for the reason why you're not getting spread reconfigured properly so that you can rule out whether -T helps your situation.
--Sharon