Fully managed cloud. 30-day FREE full featured trial. Start Now
cancel
Showing results for 
Search instead for 
Did you mean: 

AWS EXASTORAGE DISK STATE OFFLINE AND LOCKED WAITING FOR HDD RECOVERY

ImaneJabal
Padawan

Hello everyone, my name is Imane, pleased to meet you all !

I am new with using EXASOL and facing an issue with Disks and AWS while loading data to EXASTORAGE

What I am trying to do : Transfer data from AWS servers to EXASOL Database

What errors I get :

1. Exasolution - Emergency Shutdown - Reason : A volume has been locked. Shutdown system down

2. Exasolution - database corrupted - Manual intervention is needed.

A call with EXASOL SUPPORT teams identified the issue being a DISK STATE OFFLINE but we cannot identify WHY

I tried to restart the disk but it failed with an error message : Got error during EXAStorage call : Failed to enable device: operation failed

What do you suggest investigating to solve this kind of issues ?

PS: I have attached 3 snapshots to help you see error logs. hopefully it helps !

Thanks in advance

 

Imane

3 REPLIES 3

mwellbro
Xpert

Hi Imane,

just so I can wrap my head around this:

1) You set up a working exasol cluster on AWS
2) Then "while loading data to EXASTORAGE" the system failed and is currently unable to start because the disks are marked offline
3) Starting the EXAStorageService yields what you posted in Snap3

Is this correct so far ?
How did you move data "to EXASTORAGE" ? I´d expect something like "IMPORT INTO ... FROM ... " on a "SQL level" , or did you try to do something to the disks/volumes directly ?


Cheers,
Malte

ImaneJabal
Padawan

Hi Malte,

Yes correct, and regarding data transfer,i used sql query "IMPORT INTO ..."

Imane

mwellbro
Xpert

Hi @ImaneJabal ,

sorry for the delay, had a bit of a covid situation going on but seems to be all better now - since I didn´t find the time to create a test case for this scenario the following is pretty much just a guess, but maybe , since you are using 2 data disks for the volume , something happened at the "infrastructure level" ?
Did you have a look at the cloudwatch logs for this failure ?

Did exa support give you any additional pointers ?

Cheers,
Malte