LaoTsao's Weblog (老曹的網路記)

ZFS, /etc/zfs/zpool.cache and suncluster

ZFS still a very young FS and many new features with every new Solaris update and ZFS patch

This blog talk about /etc/zfs/zpool.cache behind the sence of zpool
The need of /etc/zfs/zpool.cache to speed up the import of ZFS may impact the HA-ZFS with Suncluster, the ZFS import happen before the Suncluster come to play.
In the beginning ZFS was designed to panic a system in the event of a catastrophic
write failure to a pool, since then zfs has introduce the failmode. PSARC 2007/567

The default behavior will be to "wait" for manual intervention before
allowing any further I/O attempts. Any I/O that was already queued would
remain in memory until the condition is resolved. This error condition can
be cleared by using the 'zpool clear' subcommand, which will attempt to resume
any queued I/Os.

The "continue" mode returns EIO to any new write request but attempts to
satisfy reads. Any write I/Os that were already in-flight at the time
of the failure will be queued and maybe resumed using 'zpool clear'.

Finally, the "panic" mode provides the existing behavior that was explained
above.

The syntax for setting the pool property utilizes the "set" subcommand defined
in PSARC 2006/577:

# zpool set failmode=continue pool
# zpool create -o failmode=continue pool

ZFs and suncluster

Sun alert Solution 245626 : ZFS Pool Corruption May Occur With Sun Cluster 3.2 Running Solaris 10 with patch 137137-09 or 137138-09

Resolution

This issue is addressed in the following releases:

SPARC Platform:

* Solaris 10 with patch 139579-02 or later obsoleted by 139555-08

x86 Platform:

* Solaris 10 with patch 139580-02 or later obsoleted by 139556-08

To avoid this problem, install Solaris 10 patch 139579-02 (for SPARC) or 139580-02 (for x86) immediately after you install 137137-09 or 137138-09 but before you reboot the cluster nodes.

The latest Suncluster 3.2u2 patch for solaris 10
sparc 126106-30
x86 126107-30

There are some IDR patch for ZFS performance

To boot the system without import any zpool
boot -m milestone=none
return to normal
svcadm milestone all

Posted at 08:10PM May 17, 2009 by hstsao in General  |  Comments[0]

PCA- Patch Check advanced

Analyze, download and install patches for Sun Solaris.
By chance, I find this patching tool PCA
Created by
Martin Paul
Institute of Scientific Computing
Nordbergstrasse 15/C/3
1090 Wien
Austria

It seems has many follower
Other say about pca

Patch Check Advanced (pca) generates lists of installed and missing patches for Sun Solaris systems and optionally downloads patches. It resolves dependencies between patches and installs them in correct order. It can be the only tool you ever need for patch management, be it on a single machine or a complete network. Just one perl script, it doesn't need compilation nor installation, and it doesn't need root permissions to run. It works on all versions of Solaris, both SPARC and x86.

Features

• Easily understandable and configurable format for the patch report, containing Recommended/Security status and age of a patch.


• Shows all missing Recommended/Security patches in one concise list. Only patches for packages which are actually installed are listed. Obsolete/Bad patches are ignored. Output can be formatted in HTML, with links to patch READMEs and downloads (Example).


• It analyzes the patch dependencies, and lists required patches in the correct order for installation.


• If requested, it downloads patches from Sun's patch server and installs them. One patch, groups of patches, or all missing patches. Start it, let it run, and return to a fully patched system.


• Set up a local patch server and speed up downloads tremendously.


• It's fast: Generating a complete patch report takes just a few seconds.


• It's small: One file, ca. 4000 lines, both code and documentation. Makes understanding and modifying the code for your own needs easy.


• It can assist in staying informed about firmware and other unbundled patches.


• All the information about a machine needed for analysis can be read from files, so you can use pca even if it doesn't run on the target machine.


• There's an auto update mechanism to keep pca itself up-to-date.

License

Usage of pca is free of charge for private, educational and commercial use. No responsiblity is taken for any damage caused by using pca. You may modify pca's source code to fit your local needs. If sharing modified versions of pca with others, keep a reference to the original author and distribution site.

This is a discussion and support list
To subscribe, send an empty message to pca-join@lists.univie.ac.at. To leave the list, send an empty message to pca-leave@lists.univie.ac.at.

To post to the list, send your message to pca@lists.univie.ac.at. Messages to the list from non-subscribers are allowed, but moderated. If you are not subscribed, include a note that you'd like to receive a Cc: of any reply.

Please try it out

Installation

pca doesn't need any complicated compilation, installation or registration procedure, nor root permission. It's just one perl script.

• You need perl to run pca. If you want to use any of pca's download functions, you need wget (?v1.7). Both are included in recent versions of Solaris.
• Download the script:
  
  
  
  1. stable: pca (20090408-01, Changes, Usage)
  
  
  2. develop: pca (20090506-01, Changes, Usage)
  
  
  
  and make it executable (chmod +x pca). Move it to a directory in your PATH.

  Alternatively, pca is available as an SVR4 compliant package from Blastwave (maintained by D. Clarke), from OpenCSW (CSWpca, maintained by D. Michelsen) and on Sunfreeware (maintained by S. Christensen).

• To download patches or patch READMEs from Sun, a Sun Online Account (SOA) is required. If you don't have one yet, get a free SOA and use the askauth or the user and passwd options to feed the SOA data to pca. A free SOA will grant access to security and driver patches only. To access all patches, you need to buy a Sun Service Plan and connect it to your SOA.


• Run it: pca. There is no need to run pca as root for basic usage.


• Documentation and release notes are included in the script; view it with pca --man. If you prefer documentation in man page format, get pca.8 and move it to a directory in your MANPATH.



If you are forced to use a proxy for web access, make sure that wget is configured to use it: Set http_proxy in /etc/wgetrc or $HOME/.wgetrc or use the wgetproxy option with pca.

If you do not have wget installed on your system, download the current patch cross-reference file patchdiag.xref and move it to /var/tmp/ before running pca.

Posted at 06:28PM May 17, 2009 by hstsao in Solaris 10  |  Comments[0]

xvm Ops Center

Get the Sun xVM Ops Center 2.1

Select the level of management requirements you need from one of the following offerings:

• xVM Ops Center Controller Provides the ability to access data from all proxies


• Standard Agent Pack For basic monitoring and firmware provisioning


• Premium Agent Pack For all hardware and operating system management capabilities

Sun xVM Ops Center can be purchased in two packs

• Standard Agent Pack:
  
  • Used for Monitoring and firmware provisioning when operating system management is not needed
  
  
  • Features:- Agent,- Bare Metal Discovery,- Inventory,- Firmware Update,- Monitoring


• Premium Agent Pack:
  
  • Used for Enterprise automation for customers who want support from Dock to Datacenter
  
  
  • Features:- Agent,- Bare Metal Discovery,- Inventory,- Firmware Update,- Monitoring,- Update Operating System,- Provisioning Operating System,- Profile Based Provisioning


• Premium Services: 24x7 Support Services


• xVM Ops Center Controller: - Available as a separately licensed product, - Licensed on per-server basis


• Proxy: Included with xVM Ops Center Controller Subscription


• Installation Services :Included with xVM Ops Center Controller Subscription

Operating System support

• Solaris:Solaris SPARC 8/9/10, Solaris x64/x86 9/10


• Red Hat Enterprise Linux Advanced Server:Red Hat RHEL AS/ES/WS 3/4/5


• SUSE: Novell SUSE & SLES 8/9/10

Hrdware Supported

• Sun Fire V125, V210, V215, V240, V245, V440, V445, V490, V890


• Sun Fire T1000, T2000


• Sun Fire T5120, T5220


• Sun Fire V20z, V40z


• Sun Fire X2100, X2100 M2, X2200 M2


• Sun Fire X4100, X4100 M2, X4150, X4200, X4200 M2


• Sun Fire X4450, X4500, X4600, X4600 M2


• Sun Netra 240, 440


• Sun Netra X4200 M2


• Sun Blade 6000, 6048, 8000


• Sun Blade T6320


• Sun Blade X6220, X6250


• Sun Blade X8420, X8440

Standards HW supported: Sun xVM Ops Center supports ALOM, ILOM, ELOM, and RSC service processor enabled systems.

xVM Ops Center Controller Minimum Specification

• Memory 4GB Available RAM.


• Hard Disk 1GB free in OS partitions 10GB per update channel 3GB per OS image


• Operating System Solaris 10 update 3 and above, RHEL5 Server.


• Processor AMD Opteron and Intel Xeon: 2 sockets UltraSPARC T1/T2: 1 socket; 2 or more cores UltraSPARC IV+/IV: 2 sockets UltraSPARC IIIi: 2 sockets


• Network Connection At least one Network Interface Card (NIC).

Proxy Controller Minimum Specification

• Memory 2GB Available RAM.


• Hard Disk 1GB free in OS partitions 10GB per update channel 3GB per OS image


• Operating System Solaris 10 update 3 and above, RHEL5 Server.


• Processor AMD Opteron and Intel Xeon: 1 sockets; 2 or more cores, UltraSPARC T1/T2: 1 socket; 1 or more cores, UltraSPARC IV+/IV: 1 sockets, UltraSPARC IIIi: 1 sockets


• Network Connection At least one Network Interface Card (NIC).

Due to the complexity of the xVM Ops Center, this is few software that customer can not just download and try it out:-(

Posted at 05:04PM May 17, 2009 by hstsao in xvm-V12N  |  Comments[0]

xvm-server latest

This is the latest development in xVM-server

posted by duncanha on April 30, 2009

Sun are concentrating on including xVM Server EA as a Beta-level feature of the next Ops Center release and will be pushing towards full support of the xVM hypervisor that's included with OpenSolaris in a future release of Ops Center.

The xVM Server Early Access code in Ops Center can be deployed through Ops Center only, not as a standalone product.

We have closed the separate xVM Server Early Access program because xVM Server EA is included as a feature of Ops Center.

Existing Ops Center customers who wish to upgrade to 2.1 and access the xVM Server functionality should contact their Sun sales or services representative

Posted at 04:24PM May 17, 2009 by hstsao in xvm-V12N  |  Comments[0]

Sun Storage 7xxx Unfied Storage System

Sun announced Sun Storage 7xxx Unified System the first product of the family 7110, one can find the docs here http://docs.sun.com/app/docs/coll/stor-7110

Product Description The Sun Storage 7000 Unified Storage products provide efficient file and block data services to clients over a network, and a rich set of data services that can be applied to the data stored on the system. The Unified Storage Systems include support for a variety of industry-standard client protocols, including: * CIFS * NFS * HTTP * WebDAV * iSCSI * FTP Your Unified Storage System also includes new technologies to deliver the best storage price/performance and unprecedented observability of your workloads in production, including: * Analytics, a feature for dynamically observing the behavior of your system in real-time and viewing data graphically. * The Hybrid Storage Pool, composed of optional Flash-memory devices for acceleration of reads and writes, low-power, high-capacity disks, and DRAM memory, all managed transparently as a single data hierarchy. To manage the data that you export using these protocols, you can configure your Unified Storage System using our built-in collection of advanced data services, including: * RAID-Z (RAID-5 and RAID-6), Mirrored, and Striped * Snapshots, unlimited read-only and read-write, with snapshot schedules * Built-in Data Compression * Remote Replication of data for Disaster Recovery * Active-Active Clustering (in the Sun Storage 7410) for High Availability * Thin Provisioning of iSCSI LUNs * Virus Scanning and Quarantine * NDMP Backup and Restore To maximize the availability of your data in production, the Sun Storage products include a complete end-to-end architecture for data integrity, including redundancies at every level of the stack. Key features include: * Predictive Self-Healing and Diagnosis of all System FRUs: CPUs, DRAM, I/O cards, Disks, Fans, Power Supplies * ZFS End-to-End Data Checksums of all Data and Metadata, protecting data throughout the stack * RAID-6 (DP) and optional RAID-6 Across JBODs * Active-Active Clustering for High Availability * Link Aggregations and IP Multipathing for Network Failure Protection * I/O Multipathing between the Sun Storage 7410 and JBODs * Integrated Software Restart of all System Software Services * Phone-Home of Telemetry for all Software and Hardware Issues * Lights-out Management of each System for Remote Power Control and Console Access

Posted at 03:58PM May 17, 2009 by hstsao in General  |  Comments[0]

virtual iron first look

Virtual iron first look

The Virtual Iron solution consists of three components:

• VI - Center provides a central place to control and automate virtual resources. It streamlines tasks that are normally highly manual and time-intensive and significantly reduces data center costs and complexity. Run on Linux and window server


• Virtual Iron Virtualization Services are deployed automatically on bare-metal, industry-standard servers without requiring software installation or management. These features streamline data center management and reduce operational costs. Support server with Intel-V and AMD-V capability. Support Linux and window VM


• Open Source Virtualization (Xen based)

Price $799/socket

Datasheet

Feature Benefit

1. LiveProvisioning: A "zero touch" automated deployment capability that eliminates the need for physical installation or management of virtualization software on physical servers
2. Virtual Storage Management: Create virtual storage on the fly from local, fibre channel or iSCSI storage servers
3. Virtual Network Management: Create virtual switches that allow virtual machines to use standard networks or vLANs transparently
4. Virtual Machine Creation Wizard:Use the virtual machine create wizard to simply create virtual infrastructure by cloning existing templates or creating new servers
   
   
   • Up to 8 CPUs
   
   
   • Up to 32 GB RAM
   
   
   • Up to 5 NICs with unique MAC addresses
   
   
   • Scheduling priority
   
   
   • Resource throttling
   
   

5. LiveMigration: Migrate workloads from one physical server to another, for example to provide additional capacity, without any application downtime.
6. LiveCapacity: LiveCapacity optimizes virtual machine performance and data center utilization. It accomplishes this by rebalancing nodes by moving virtual machines when the node's CPU utilization is above a threshold.
7. LiveRecovery: Automates virtual machine recovery from physical hardware failures without the cost and complexity of clustering software. Virtual machines can be automatically restarted on new hardware when physical hardware fails, reducing outage duration and operational costs
8. LiveMaintenance: LiveMaintenance automates the process of taking a physical server out of service for maintenance, such as adding more physical memory, without impacting the virtual machines running on it. LiveMaintenance uses LiveMigrate to move virtual machines -- without stopping operating systems or applications -- to other nodes in the virtual data center.
9. LiveSnapshot: LiveSnapshot creates space-efficient point in time images of running or stopped virtual machines. LiveSnapshot makes it easy to try patches and roll back to previous states, or backup a running virtual machine.
10. LivePower: LivePower optimizes data center power consumption by monitoring resource utilization in the virtual data center. When there is excess CPU capacity, LivePower consolidates virtual machines onto fewer servers and shuts down physical servers that are not running virtual machines, based on pre-defined policies. When virtual machine load increases beyond pre-defined thresholds, LivePower turns on physical servers and LiveMigrates virtual machines to rebalance the virtual data center and ensure that resource requirements and service levels are met.
11. LiveConvert: An automated software solution powered by PlateSpin that enables customers to easily migrate 32-bit Windows workloads (data, applications, and operating systems) from physical, virtual and image-based infrastructures. LiveConvert provides Virtual Iron users with the ability to quickly migrate workloads between physical servers and virtual machines allowing users to quickly achieve the benefits of large-scale server consolidation.
12. Jobs and Reports: View historical information on system performance or administrative actions in the data center. Useful for capacity planning, billing, and 404 auditing.
13. User based access control: Perform user management locally or connect with Microsoft Active Directory or LDAP servers. Prevent unauthorized users from using the system. Monitor changes based on user actions for compliance purposes.

Posted at 03:57PM May 17, 2009 by hstsao in xvm-V12N  |  Comments[0]

Sun Net Connect to be disabled on May 30, 2009

Email send to all registered Net Connect customer on march 11, 2009

Sun Alert 257308 "On may 30, 2009 Sun Net Connect Service will be disabled. Failure to uninstall Client may cause downtime"

replacement solutions

1. Services Tools Bundle (STB)


2. Sun Secure File Transport (SFT)


3. Shared Shell


4. xVM OpsCenter


5. Auto Service Request (ASR) ASR require STB the download page


6. Sun Inventory Channel



Posted at 11:34AM May 17, 2009 by hstsao in General  |  Comments[0]

solaris 10 05/09 u7 release what's new

what new features of 05/09 (u7) release:
what new

• Support Added for Using ZFS ClonesWhen Cloning a Zone
  
  
  
  1. If the source and the target zonepaths reside on ZFS and both are in the same pool, a snapshot of
     the source zonepath is taken and the zoneadm clone uses ZFS to clone the zone.
  
  
  2. You can specify to copy a ZFS zonepath instead of specifying to clone the ZFS. If neither the
     source nor the target zonepath is on ZFS, or if one is on ZFS and the other is not on ZFS, the
     clone process uses the existing copy technique.
  
  

• iSCSI target update, performance and reliability improvements
  
  
  
  1. Improved TCP/IP timeout recovery
  
  
  2. iSCSI initiator invoked SCSI RESETs
  
  
  3. Code path and memory leak cleanup
  
  
  4. Improved interoperability with Target Port Group Tags (TPGT), unidirectional and
     bidirectional CHAP authentication, and RADIUS server support
  
  
  5. Improved Internet StorageName Service (iSNS) support, including recovery from
     unavailable iSNS servers
  
  
  6. Updated SCSI-3 Persistent Reserve functionality that enables the use of the functionality in
     various clustering solutions on both Solaris and other operating systems
  
  
  
  
• The Solaris iSCSI Target release now supports a wide variety of iSCSI initiators for the following
  operating systems:
  
  
  1. Solaris 10
  
  
  2. OpenSolaris
  
  
  3. Linux: RedHat Enterprise Linux (RHEL), Suse, and Ubuntu
  
  
  4. VMWare ESX
  
  
  5. Microsoft Windows (XP, Vista, Server 2003, Server 2008, Windows Cluster Server)
  
  
  6. MacOSX
  
  
  
  

Posted at 10:37AM May 17, 2009 by hstsao in Solaris 10  |  Comments[0]

QDR Magnum infiniband Switch

During the sc08 Sun did showcase the new QDR Magnum infiniband Switch
QDR Magnum infibiband Switch

• Based on Mellanox InfiniScale IV 36-port 40 Gb/s Infiniband Switch Silicon


• 36 4X or 12 8X or 12 12X InfiniBand ports, or a combination of port types


• PCI Express 2.0 x4 5GT/s (1.1 compatible)


• three stages CLOS


• 648 4x port presented in 216 12x CXP connectors


• Host based Subnet manager

The SB6048 also introduce QDR NEM QNEM

• 30 ports of 4x QDR InfiniBand and 24 ports of GbE


• Up to 6:1 cable reduction as compared to competitive systems


• Scalability up to 5,000+ node clusters

Sun PCIe Dual port QDR IB HCA and DDR FEM
QDR HCA, DDR FEM

General observation on QNEM:

• Function as leaf switch


• bases for 3 D torus IB Fibre


• with other core switch one can build 3 stages or 5 stages or 7 stages IB fibre etc

Posted at 09:59AM May 17, 2009 by hstsao in Infiniband  |  Comments[0]

ZFS

One of best feature in Solaris 10 is ZFS

ZFS was introduced in Solaris 10 update 2 , but ZFS is very young FS, there are many advancement in opensolaris and some are backported in Soalris 10.

Recently the intruducation of Openstorage and Unified Storage Appliance future introduce features that only presented in Unified Storage Appliance.

• ZFS Configuration Guide
• ZFS Best Practice Guide
• ZFS Evil Tuning Guide
• ZFS Troubleshooting Guide
• ZFS for Database

IMHO, Sun should speedup putting features/fixes back in Solaris 10, so many enterprise customer can take advantage of ZFS features.

Key features:

• user quota, right now one need separate FS for user quota and become issues when one need to setup quota for thousands users

Posted at 09:18AM May 17, 2009 by hstsao in Solaris 10  |  Comments[0]

coolstack, webstack and solaris zone

This is concern on AMP and solaris zone

coolstack 1.3.1 has version on sparc and x86, webstack also has version on sparc and x86

coolstack install pkg under /opt/coolstack

webstack install pkg under /opt/webstack

opensolaris 2008.05 only support x86, next release will support sparc.

the amp, amp-dev install apache(apache2) under /usr, MySQl under /usr/mysql and php under /usr/php etc

the blastware.org also provide amp and install pkg under /opt/CSW

As for AMP in solaris zone, the configuration file and some docs files are under /etc and /var so it is possible to use the spare zone and setup different zone with various different zone for various AMP.

Even under /opt , one can share /opt with various zone

Please see the following links:

coolstack 1.3.1
web stack 1.4

please see this link for MySql in solaris containers

some example of apache in zone

Posted at 07:50AM May 17, 2009 by hstsao in Solaris 10  |  Comments[0]