Contents
1 Introduction
If an MTAS-related process terminates in an unexpected way, an error dump file is generated, and, if needed, the terminated MTAS process is automatically restarted. There are two kinds of MTAS processes: vDicos processes and Linux processes. Most MTAS processes execute as vDicos processes, and if a vDicos process ends for some reason (assertion failure, segmentation fault, and so on), this is referred to as a "Capsule Abortion". Other MTAS processes execute as Linux processes, and if a Linux process ends, this is referred to as a "core dump".
Whenever a Capsule Abortion or a core dump occurs, an error dump file is generated and the related log files are collected for troubleshooting purposes.
The error dump files, along with the collected log files, are put into an error dump archive file (tgz-file), which may thus contain one or several error dump files.
Whatever caused the error dump can also have generated one or several alarms. If the reason for the error dump cannot be found, the error dump file archive should be collected and sent to the next level of maintenance support for further analysis.
1.1 Prerequisites
This section describes the possible documents, tools, and required conditions before starting the data collection procedure.
It is expected that the user has prior knowledge about data and telecommunication, including knowledge about the MTAS and the virtualized environment. It is assumed that the user is familiar with concepts, terminology, and abbreviations within these areas.
1.1.1 Tools
The following tool must be available before performing any procedure in this document:
- A workstation with SSH client.
2 Supervision
The generation of error dump files and error dump archive files is monitored by the operating system. Error dump files are automatically collected, compressed, and maintained in the file space where the error dump archive files are stored. The default value for the number of stored error dump archive files is 30. The older error dump archive files are removed if the number of the error dump archive files exceeds the threshold, for more details, refer to LEM Error Dump and Log User Guide.
3 Files and Archives
The error dump files are transferred to the System Controllers (SCs) for storage to file in a dedicated directory, named the error dump archive directory.
- Directory /opt/cdclsv/storage/dumps/
The error dump archive directory is created when the first MTAS error dump file is generated. It contains compressed vDicos process abortion dump files, core dump files, and different other log files collected from different locations, into an error dump archive file. Files in the error dump archive directory use the following naming pattern: logdump-<date>-<time>.tgz.
4 Data Collection
Whatever caused the error dump can also have generated one or several alarms. If the reason for the error dump cannot be found, the error dump file archive should be collected and sent to the next level of maintenance support for further analysis.
The MTAS Data Collection feature with basic profile can be used to collect the MTAS error dump archive files, for more information about the MTAS Data Collection feature, refer to Data Collection Guideline for MTAS.