failures in installing Vertica
>> Validating node and cluster prerequisites...
Failures during local (OS) configuration for verify-15.56.9.63.xml:
HINT (S0151): https://my.vertica.com/docs/7.0.x/HTML/index.htm#cshid=S0151
These disks do not have known IO schedulers: '/dev/mapper/vg00-crash'
('') = '', '/dev/mapper/vg02-lvol01' ('') = '',
'/dev/mapper/vg01-lvol01' ('') = '', '/dev/mapper/vg00-var' ('') = '',
'/dev/mapper/vg00-usr' ('') = '', '/dev/mapper/vg00-tmp' ('') = '',
'/dev/mapper/vg00-tidal' ('') = '', '/dev/mapper/vg00-opt' ('') = '',
'/dev/mapper/vg00-home' ('') = ''
HINT (S0305): https://my.vertica.com/docs/7.0.x/HTML/index.htm#cshid=S0305
TZ is unset for verdba. Consider updating .profile or .bashrc
WARN (S0170): https://my.vertica.com/docs/7.0.x/HTML/index.htm#cshid=S0170
lvscan (LVM utility) indicates some active volumes.
FAIL (S0020): https://my.vertica.com/docs/7.0.x/HTML/index.htm#cshid=S0020
Readahead size of /dev/mapper/vg00-crash is too low for typical systems:
256 < 2048 Readahead size of /dev/mapper/vg02-lvol01 is too low for
typical systems: 256 < 2048 Readahead size of /dev/mapper/vg01-lvol01 is
too low for typical systems: 256 < 2048 Readahead size of
/dev/mapper/vg00-var is too low for typical systems: 256 < 2048
Readahead size of /dev/mapper/vg00-usr is too low for typical systems:
256 < 2048 Readahead size of /dev/mapper/vg00-tmp is too low for typical
systems: 256 < 2048 Readahead size of /dev/mapper/vg00-tidal is too low
for typical systems: 256 < 2048 Readahead size of /dev/mapper/vg00-opt
is too low for typical systems: 256 < 2048 Readahead size of
/dev/mapper/vg00-home is too low for typical systems: 256 < 2048
System prerequisites failed. Threshold = WARN
Hint: Fix above failures or use --failure-threshold
Installation FAILED with errors.
****
AdminTools and your existing Vertica databases may be unavailable.
Investigate the above warnings/errors and re-run installation.
****
[root@c0043087 vertica]#
Comments
The output that you have pasted contains a list of URLs with instructions on how to resolve each identified issue. Have you read through and followed those instructions?
The HP Cloud supports many different platforms and configurations; Vertica runs optimally with one particular configuration.
Also, please be aware that Vertica does not provide formal support for the HP Cloud at this time. There are a number of issues that you are likely to run into, largely regarding networking functionality and persistent-storage performance. (The HP Cloud offering is by default optimized for serving content; but most of Vertica's activity is intra-node, not from each separate node to the Internet, so it benefits from a different physical hardware configuration.) These can be resolved, or at least mitigated; you'll want to read through our cloud-related documentation carefully, though, to understand what's going on.
Adam
From these four failures, it seems all related to the LVM. I read through the link
WARN (S0170): https://my.vertica.com/docs/7.0.x/HTML/index.htm#cshid=S0170
lvscan (LVM utility) indicates some active volumes.
but it only states one line that "HP Vertica does not support LVM (Logical Volume Manager) on any drive where database (catalog and data) files are stored. The installer reports this issue with the identifier: S0170.". i don't know what I'm supposed to do next?
Where can i find the guidance to the HP cloud configuration?
I have 3 Cloud server nodes, and am going to install on all of them to form a cluster. hope that would work..
Again, thanks for your response.
Best regards
Johnny
With regard to LVM, that message is because you have chosen a VM image that uses LVM to host its root filesystem. The Vertica installer can't tell up front which of the available partitions you plan to install databases on (since you can create databases at any time after installation), so it will always give this warning if it sees any LVM partitions.
The easiest resolution in general in the cloud is to attach a new volume to store your data and catalog, and to manually format that volume using a simple single large partition. If that's your plan, you can safely follow the instructions in the text above regarding using "--failure-threshold" to ignore the warning.
We also do have users using Vertica on top of LVM. (There are other threads on these forums about LVM; feel free to search for them.) It's not supported because some LVM implementations, particularly on older platforms that we still support, have bugs that can cause data corruption under certain unusual circumstances. But it usually works...
Also, don't forget the timestamp warning above. That one can lead to very confusing issues if you work with timezones a lot. And the FAIL message about readahead size should be addressable for any type of partition.
Regarding HP Cloud instructions, Vertica doesn't provide specific instructions for HP's Cloud offering. Interestingly, we do provide information regarding Amazon's cloud; there are a lot of pieces to configure, but you might find the document instructive:
https://my.vertica.com/docs/7.0.x/PDF/HP_Vertica_7.0.x_Vertica_AWS.pdf
(The HP Cloud is much newer; obviously we'd like to support it, but right now I can just comment on docs that we have released.)
Adam
Here is an example of how you can use the flag:
/opt/vertica/sbin/install_vertica --hosts 10.123.1.123 --failure-threshold NONE
-the NONE value will tell the installer to ignore any pre-reqs.
-use this option only when you have WARN and HINT thresholds - if you get any FAIL's make sure you fix them before.