This topic describes how to download diagnostic bundles to use for troubleshooting an
Impala Virtual Warehouse in Cloudera Data Warehouse (CDW) Public Cloud.
To troubleshoot issues with your Impala Virtual Warehouses, download
diagnostic bundles of log files for the sidecar containers that support Impala components and
for the components themselves. These diagnostic bundles, in the form of ZIP files, are stored
in your public cloud account. Currently, this is only available on AWS environments.
Required role: DWAdmin
Before you can download log files you must run workloads on your Impala Virtual
Warehouse.
Log in to the CDP web interface and navigate to the CDW service.
In the CDW service, click Overview in the left
navigation panel, and select the Impala Virtual Warehouse for which you want to
download a diagnostic bundle.
In the Impala Virtual Warehouse tile, click the more options menu, and select
Collect Diagnostic Bundle:
The Diagnostic Bundle Options dialog box launches:
In this dialog box, you can choose or set the following options:
By Time Range: Select a specific time range of log files
to generate from the drop-down list, or you can choose a custom interval in the
next option.
By Custom Time Interval: Select the start and end time
from the drop-down list to define the specific time interval for the log files in
the diagnostic bundle.
Collect For: Remove any of the available categories of
log files by clicking the adjacent X:
Workload contains the log files for the Impala
components, such as statestored, impalad,
catalogd, coordinator, auto-scaler, and Hue frontend and
backend logs. See Impala Logs.
Sidecar contains the logs for the sidecar containers
that support Impala components.
HMS contains sidecar container logs that support the
metastore.
After selecting which logs you want to generate for the diagnostic bundle, click
Collect to generate the bundle.
After generating the diagnostic bundle, a message appears at the top of the browser
window like the following:
Collection of Diagnostic Bundle for impala-mpvt initiated. Please go to
details page for more information.
This message means that your diagnostic bundle generation job was started. Usually,
it takes about 10 seconds to finish, and then you can go to the details page for
information about downloading the bundle from a location in AWS S3. If you have a large
cluster and are collecting logs for a longer period of time, it might take longer to
generate the diagnostic bundle.
To navigate to the Virtual Warehouse details page, click Virtual
Warehouses in the left navigation menu, and then, in the list of
warehouses, select the name of the Virtual Warehouse for which you just generated the
diagnostic bundle.
In the details page for that Virtual Warehouse, click the DIAGNOSTIC
BUNDLE tab, and then click the copy-to-clipboard icon to copy the S3
bucket path to the diagnostic bundle:
After you copy the S3 bucket location path to your clipboard, paste it in a text
document and use it to navigate to the location in your Amazon S3 management console
to download the ZIP file.
When you expand the diagnostic bundle ZIP file that you downloaded from S3, there are
four directories that contain log files as described in Step 4. In addition, there is a
diagnostic-data-generator.log file, which contains useful information
about the diagnostic bundle generation that might be helpful in your troubleshooting.