Dell EqualLogic PS6210XS EqualLogic Host Integration Tools for Linux Version 1 - Page 105
Troubleshooting, General Troubleshooting Guidelines
View all Dell EqualLogic PS6210XS manuals
Add to My Manuals
Save this manual to your list of manuals |
Page 105 highlights
8 Troubleshooting This chapter provides troubleshooting information for HIT/Linux. Topics: • General Troubleshooting Guidelines • Using the eqllog Facility • Using the udev Facility • Using the ehcmcli Diagnostic Tool • Running the ehcmd Daemon • Considerations When Partitioning iSCSI Volumes General Troubleshooting Guidelines The following general troubleshooting guidelines apply when using HIT/Linux. Ensuring Correct Operation of the dm-switch Kernel Module If the dm-switch kernel module appears to be installed but fails to operate correctly, examine the /var/log/messages file because the dm-switch kernel module writes status messages using the standard kernel logging facility. The log file might reveal a problem (for example, a system resource issue). If that occurs, contact your system administrator to adjust the environment to resolve the issue. The ehcmd daemon writes debugging information into a log file in the /var/log/equallogic/ directory. Error messages in the log file might provide important clues about the interaction between the HIT/Linux components and the operating system environment. For example, if Device Mapper commands (that is, the dmsetup commands) are failing, the commands are written in the log along with the errors returned by Linux. The errors might point to environmental issues that need to be resolved by the system administrator before multipath I/O can operate at peak efficiency. Establishing iSCSI Sessions If the ehcmd daemon is having trouble establishing the desired number of sessions, look at the event log on the PS Series group. In particular, you can look at the existing sessions and access control lists on the volume and then verify that all the Ethernet ports are allowed to access the volume if IP ACLs are being used. If multiple Linux initiators are connecting to the same volume, set the option Allow simultaneous connections from initiators with different IQN names. NOTE: To prevent data corruption, set this option only when you have a cluster-aware file system or application using the volumes. If the ehcmd daemon is still unable to connect from all Ethernet ports, a host or networking issue might be causing the problem. To narrow the scope of the problem, use iscsiadm to manually attempt to log in to a volume from each Ethernet port. In addition, you can run eqltune and look at diagnostic information to configure an optimal Linux multipath I/O environment. For information on using eqltune, see EqualLogic Host Performance and Tuning Suite (eqltune). Increasing the Volume Size on a PS Series Group You can increase the size of a volume on the PS Series group. The ehcmd daemon periodically rescans all devices for the volumes that it manages for multipath I/O. If ehcmd detects the change in volume size, it attempts to rebuild the two-tier device configuration using the newer, larger capacity. Troubleshooting 105