Release notes for the com.oonics Enterprise IT Platform Version 5.0
releasenotes description
Copyright © 2011 ATIX AG
Copyright © 2011 ATIX Informationstechnologie und Consulting AG, called “ATIX” within this document.
The text of and illustrations in this document are licensed by ATIX AG under a Creative Commons Attribution–Share Alike 3.0 Unported license ("CC-BY-SA"). An explanation of CC-BY-SA is available at http://creativecommons.org/licenses/by-sa/3.0/. In accordance with CC-BY-SA, if you distribute this document or an adaptation of it, you must provide the URL for the original version.
ATIX AG, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section 4d of CC-BY-SA to the fullest extent permitted by applicable law.
Red Hat, Red Hat Enterprise Linux are trademarks of Red Hat, Inc., registered in the United States and other countries.
Linux® is the registered trademark of Linus Torvalds in the United States and other countries.
All other trademarks are the property of their respective owners.
The Release Notes provide high level coverage of the improvements and additions that have been implemented in the current com.oonics Enterprise IT Platform.
For detailed documentation on all changes to the com.oonics Enterprise IT Platform update, refer to the ATIX Bugzilla.
This part is intended to give an overview of features from a economical and technical perspective without going into details.
The com.oonics Enterprise IT Platform consolidates server and storage systems. It enables high availability, optimal scalability and ease of management. The com.oonics Enterprise IT Platform is suitable for companies of any size and can be adapted to meet the individual user's needs. In time, the needs evolve. It is crucial to keep IT infrastructure modular and flexible, so that is easy to extend, upgrade or re-purpose. The com.oonics Enterprise IT Platform consists of flexible modules, industry-standard components based on Linux. Due to its open source nature, the Platform allows freedom, flexibility and ease of change. Furthermore, the com.oonics Enterprise IT Platform solves the main challenges of modern IT – it optimizes the performance and availability of IT services and applications, such as file and web services, databases (Oracle, PostgreSQL, MaxDB, MySQL etc.), advanced ERP systems such as SAP and server virtualization. com.oonics is also a way to keep maintenance costs and TCO low.
The com.oonics modular IT infrastructure design may be depicted as follows:

Each server in the IT environment has direct access to SAN storage system. Servers may read and write on the shared storage concurrently. A shared file system enables parallel data access for all application servers. com.oonics allows servers to provide services as a cluster, without need of separate, local hard disks. Not only data and configuration are centralized but com.oonics Enterprise IT Platform allows the sharing of the operating system among cluster nodes. The server hardware is logically assigned to clusters. The assignment is not static, it may be changed in cluster runtime, without downtime. For example, some database cluster servers (blue) may be reassigned and become web servers (yellow), without the need of costly downtime or maintenance window. This increases the manageability of the systems and allows flexible changes, e.g. in case of load peaks. Since the cluster installation resides on SAN, it is easier to roll out and deploy new environment for production, tests or development. As long as the clustering and application software support it, a rolling update is significantly simplified.
- 413: com-rmcdsl fails with CDSL not existant if one or more node cdsls do not exist
No release notes found.
- 406: Change all commands to be using defaultfiles as defaults for commandline options
All commands of this product will now be able to use Property configuration files for specifying defaults without the need to give those options at command line. This is extremely useful if commands are executed very often and have a set of options that defer from the defaults. In this case you just need to create the configuration file with those defaults and don't need to specify the defaults any more.
- 426: com-dsh does not understand hostnames with 2 or more - in the name
The -R / --norange option will selectively switch of hostname range interpretation so that hostnames that appear to look like ranges will be accepted.
- 407: Change all commands to be using defaultfiles as defaults for commandline options
All commands of this product will now be able to use Property configuration files for specifying defaults without the need to give those options at command line. This is extremely useful if commands are executed very often and have a set of options that defer from the defaults. In this case you just need to create the configuration file with those defaults and don't need to specify the defaults any more.
- 435: com-dsh default terminal window size
It is now possible to use com-dsh within other scripting environments. The -T/-t option now supports this.
- 426: com-dsh does not understand hostnames with 2 or more - in the name
The -R / --norange option will selectively switch of hostname range interpretation so that hostnames that appear to look like ranges will be accepted.
- 263: com-sysreport: cannot execute sysreport and will fail
com-sysreport: did not work because no templates are available. There is an rpm comoonics-cs-sysreport-templates that fixes this problem.
- 368: com-mkcdsl on SLES11: RuntimeError: maximum recursion depth exceeded while calling a Python object
No release notes found.
- 436: Support for com.oonics enterprise copy and ext4 filesystem
The com.oonics enterprise copy suite now supports EXT4 as root filesystem.
- 64: [RFE] ComFilesystemCopyset: Excludelist should be added
It is now possible to add exclude lists to ComFilesystemCopysets. This leads to content not been copied during the filesystem clone (e.g. /var/crash, /var/tmp, /var/run/*).
- 408: Change all commands to be using defaultfiles as defaults for commandline options
All commands of this product will now be able to use Property configuration files for specifying defaults without the need to give those options at command line. This is extremely useful if commands are executed very often and have a set of options that defer from the defaults. In this case you just need to create the configuration file with those defaults and don't need to specify the defaults any more.
- 358: Initramfs consumes more and more during runtime. Which can lead to no free memory
No release notes found.
- 183: [RFE] Progress element during comoonics boot ("Trüllerli")
The progress bar during boot of initrd is now supported via the rpm comoonics-bootimage-listfiles-plymouth and will only run on systems supporting plymouth.
- 245: Add generic network options to cluster.conf
No release notes found.
- 400: Feature request - numbered/ordered steps in com-step mode
Inside the initrd console or chroot environment listBreakpoints now list all breakpoints as they are called not ordered.
- 402: Trigger mkinitrd when the kernel is updated
com.oonics bootimage packages now integrate into the Red Hat scripts to automatically create new initrds during kernel update.
- 50: [RFE] fenceack server hangs in shell
The fenceackserver now allows running as sshd daemon.
- 431: Systems with multiple different nics get mixed up the network configuration while booting
The automatic detection of the network interfaces has been rewritten/modified so that only the relevant network interface will be configured and will be stabilized during multiple reboots.
- 356: Device changes not applied in chroot environment when chroot on local disk
No release notes found.
- 357: NFS4 OSR Cluster fails to boot on RHEL5 - wrong mount options and no portmap
No release notes found.
- 405: multiple post/pre.mkinitrd.d files will not be executed only the first one
Now more then only the first dependent script will be executed either during initrd build or during bootsr.
- 411: Fenceacksv rename to imsd (integrated management server device)
The fenceacksv product has been renamed to imsd (integrated management server device).
- 412: Use of cachefile to speed up creation of initrd
In order to speed up the creation of an initrd the cachefiles to store the files being copied will be used. Those files will be updated whenever a new rpm is released. The cachefiles can be found in /var/cache/comoonics-bootimage
- 418: Only use /usr/comoonics/sbin/init if apropriate (chroot needed)
The modified init binary is only used where needed. This means all clusterfilesystems take use of the advantages provided by the modified init binary. All local filesystems and nfs don't need it so they are not using it.
- 441: comoonics-bootimage-extras-flexd will not create /dev/flexd* devices when used with device-mapper-multipath
The comoonics-bootimage-extras-flexd package provides persistent disk device naming over different hardware/virtualization platforms. It will lead to devices called /dev/flexd[a-z][a-z]* that can identically be used with the same machine under different hardware platforms like e.g. bare metal, kvm, xen, vmware, .. .
- 367: [RFE] Bridged interfaces do not support additional attributes when used as boot interface
This feature implements properties being specified for every NIC configuration. This allows to add userspecified options to NIC configurations (e.g. BONDING_OPTS, DELAY, ..).
- 372: Feature request - fence_virsh agent support in chroot environment
RPM for listfiles for the virsh fenceagent
- 382: "Superblock last write time in the future" by boot process
Comoonics 4.6 will not set the hwclock on systems where this is not required. This lead on some systems to a unnecessary filesystem check.
- 399: shutdown -r does a power off
With /var/run mounted as cdsl a specific runlevel detection for detecting reboot type has been implemented to work.
- 431: Systems with multiple different nics get mixed up the network configuration while booting
The automatic detection of the network interfaces has been rewritten/modified so that only the relevant network interface will be configured and will be stabilized during multiple reboots.
- 259: manage_chroot -a does not understand @map tag.
The listfiles now understand all syntax independently from if they are used in chroot or initrd.
- 365: The bootprocess hangs foreever while unloading modules
No release notes found.
- 378: Mountopts in cluster.conf does not work with nfs
No release notes found.
- 383: SLES10 NFS portmap will not start in initrd
No release notes found.
- 384: SLES10 repository is not signed
Now the repomd.xml is also signed with the ATIX Key.
- 405: multiple post/pre.mkinitrd.d files will not be executed only the first one
Now more then only the first dependent script will be executed either during initrd build or during bootsr.
- 419: Only use /usr/comoonics/sbin/init if apropriate (chroot needed)
The modified init binary is only used where needed. This means all clusterfilesystems take use of the advantages provided by the modified init binary. All local filesystems and nfs don't need it so they are not using it.
- 441: comoonics-bootimage-extras-flexd will not create /dev/flexd* devices when used with device-mapper-multipath
The comoonics-bootimage-extras-flexd package provides persistent disk device naming over different hardware/virtualization platforms. It will lead to devices called /dev/flexd[a-z][a-z]* that can identically be used with the same machine under different hardware platforms like e.g. bare metal, kvm, xen, vmware, .. .
- 335: RHEL 5U2 does not provide -m option in fence_tool | joining fencedomain always fails
Check for fence_tool availability of "-m" option has been added.
- 210: permissions for chroot environment are not sufficient
No release notes found.
- 251: ccsd communication errors -FIX INSIDE-
No release notes found.
- 272: Cluster IPAddress will not show up when booting SLES10 without dhcp
fixed bug in network-lib.sh of sles10
- 358: Initramfs consumes more and more during runtime. Which can lead to no free memory
No release notes found.
- 23: [RFE] Filesystemcheck of GFS-Filesystems after a crash
PREVIEW: rootfsck can be given as bootparameter and will cause a filesystemcheck on the rootfilesystem.
- 276: Disabling NIC power up in initrd
Added a onboot="no" attribute to eth configuration. This will disable the initrd to power on the NIC. This might only make sense when booting up local filesystems.
- 330: [RFE] Specify compression tool for mkinitrd
There is a variable for mkinitrd specified called compression_cmd and compression_opts that is by default gzip but might be overwritten in /etc/comoonics/bootimage/comoonics-bootimage.cfg.
- 50: [RFE] fenceack server hangs in shell
The fenceackserver now allows running as sshd daemon.
- 242: New getParameter functionality returns wrong results if used with default parameters
bootimage: New getParameter functionality returns wrong results if used with default parameters. Fixed will return defaults.
- 267: SCSI-Failover: RDAC dependencies missing
SCSI-Failover RDAC: added dependencies that would have to be added manually to have it working.
- 341: No usb keyboard available during comoonics boot process on IBM System x3650
No release notes found.
- 266: bootsr initscript does not reload and restart as expected
bootsr: initscript supports reload and restart options
- 280: manage_chroot.sh does not output errormessages to stderr but to stdout
With no chroot bootsr will fail and create error files in /var/run...
- 312: [devicemapper-multipath] kpartx is missing in initrd
In the OSR initrd rootfilesystems or LVM logical volumes will now be found if residing on partitions.
- 314: [usability 1.1] Cluster configuration missconfigured
If the cluster configuration is missconfigured (validation fails) it is now reported during the build of the initrd and during boot of a clusternode.
- 324: IGMP v2 is not default with RHEL5U2 and above
For RHEL5 cluster sometimes it is important to fix IGMP Version to V2. This is now possible.
- 346: Oracle Enterprise Linux could not be detected as Red Hat Clone
Oracle Enterprise Linux can now be detected correctly as Red Hat Clone so that the bootscripts are working.
- 347: The mkinitrd tool does not accept the -M option for specifying additional modules
mkinitrd will now correctly accept the -M option for specifying additional modules.
- 356: Device changes not applied in chroot environment when chroot on local disk
No release notes found.
- 357: NFS4 OSR Cluster fails to boot on RHEL5 - wrong mount options and no portmap
No release notes found.
- 170: initrd fails with dm_multipath in RHEL5
Support for device-mapper multipath added
- 231: Fence_tool option -c is depricated in RHEL5U2
Updated current version to not use fence_tool -c but use it without.
- 179: xen guest will not be detected with rhel4
XEN guest detection runs with RHEL4 guests.
- 193: It happens, that the cluster gets stuck in activating volume groups
It can happen, that the cluster stucks inside the initrd when trying to activate lvm volumegroups after cman has been started. This happens when 3 of nodes are booted at the same time, i.e. without a quorum disk.
- 210: permissions for chroot environment are not sufficient
No release notes found.
- 251: ccsd communication errors -FIX INSIDE-
No release notes found.
- 156: <scsi failover="mapper"/> in cluster.conf does not activate multipath configuration
No release notes found.
- 209: initrd will not boot when locking_type=2 in /etc/lvm/lvm.conf
No release notes found.
- 267: SCSI-Failover: RDAC dependencies missing
SCSI-Failover RDAC: added dependencies that would have to be added manually to have it working.
- 241: manage_chroot.sh -a start_service does not closed fds of redirected io
No release notes found.
- 262: Plugins will not work in fenceacksv
Fix for support of plugins in fenceacksv
- 265: fenceacksv initscript will not stop fenceacksv if called with stop parameter
Fixed bug so that fenceacksv can be stopped by initscript
- 313: fenceacksv fails because /usr/bin/logger has been moved in RHEL5U3b
No release notes found.