Could n't able to restart vertica database

Are you trying to restart the entire DB or just a node in the DB?

I am trying to restart the entire vertica database

Then try using the start_db admintool.

admintools -t start_db -d <<dbname>>

It was giving me the below issue "Error: Shutdown: aborting shutdown"

What do you get from this command?

admintools -t list_allnodes

Node | Host | State | Version | DB
-----------------+---------------+-------+-------------------+------
v_pmdb_node0001 | 172.27.21.249 | UP | vertica-7.1.2.000 | pmdb

Do you have a single node cluster? If so, looks like it is up. Can you connect to the pmdb database using vsql?

Yes We have a single node cluster. We can also connect using vsql

I'm confused. So you want to restart the database that is up? Why?

Anyway, if that's what you want to do, do this:

admintools -t stop_db -d pmdb
admintools -t start_db -d pmdb

I have already tried to stop and then start it. Still facing the above issue which saying Error: Shutdown.

We are trying to restart it as there was an issue with the report extraction in OBR.
The issue is as below:
"The following database error occured: [Vertica]VJDBC ERROR FileColumnReader block /vertica/datafiles/pmdb/v_pmdb_node0001_data/129/45035999921377129/45035999921377129_0.fdb @ 0's CRC 35ae46f1f9e0c65a doesn't match record 2203c0ee4b544bc2."

Oh. Ok.

See:
https://forum.vertica.com/discussion/207463/strange-error-message-during-query

We are not able to start the vertica databse now with the above command. Can you please help.

Even after restoring the vertica database, still not able to start the database. Can you please help ?

What does this command return?

admintools -t return_epoch -d pmdb

LastEpoch|0

Hi Jim, Can we discuss over a call if you have time ?

Hi Jim, We have used the below command to start the vertica db.
admintools -t start_db -d pmdb -F

However It's not getting completely started ie in the list 2 status is INITIALIZING and rest of the node status is DOWN.

Can you please help.

@sprusty - Did you get the node started?

If not, try the attached script to manually start the node. I had to name the file with the .txt extension to upload it. Rename it to manual_start_up.sh when you DL it.

Copy the file to your node and make sure it is executable. Then run it like this:

./manual_start_up.sh pmdb

I'd also try renaming the corrupt file to something else:

mv /vertica/datafiles/pmdb/v_pmdb_node0001_data/129/45035999921377129/45035999921377129_0.fdb /vertica/datafiles/pmdb/v_pmdb_node0001_data/129/45035999921377129/45035999921377129_0.fdb_old

Then try to start the DB:

admintools -t start_db -d pmdb -F

Normaly Vertica will recover the file, but since you only have one node that's not going to happen. So you might lose some data.

Hi Jim, During running the script it was giving me an error saying "/bin/bash^M: bad interpreter: No such file or directory". So I ran a command sed -i -e 's/\r$//' scriptname.sh and then executed the same. During executing the script it said, database is up, however when checking it was showing DOWN.

INFO: Database pmdb is not running

Kindly help

@sprusty - Did you try renaming that file?

Thanks Jim for your continuous response on this.
Currently the DB is up after setting the projections to 0.

What should be our next course of action.
Regards
Suryakanta