I experienced the following error on starting up a GPFS node in a cluster: unexpected token conflict in recovery: majType 1 minType 7 tokType Inode key F657ED5089ABDE89:00000000000FC191:0000000000000000 node 1 mode xw flags 0x0 seqNum 1141739 The resulting in the node in question asserting and other nodes remote asserting, making the cluster unstable. IBM states this is related to the token memory (TM) management which can lead to unexpected result when being exhausted....

I had a need to run my home router as a router-on-a-stick, i.e. multiple VLANs on a single interface. This requires a switch with VLAN support. I already have a Cisco E4200 as a 5Ghz WAP with 1Gb switches which I would like to use. Using the DD-WRT firmware, it supports VLAN tagging. The rest of this post is my experience getting VLAN tagging to work. Objective The objective was to use all 5 ports as switch ports (the 4 LAN ports and the “Internet” port)....

At the University of Bristol, we are currently in the process of migrating from one storage vendor to another for our block storage requirements. This includes our Oracle databases running on Solaris 10. In an attempt to minimise downtime, a number of other changes were scheduled at the same time, including moving from Fibre Channel to iSCSI and moving file system from UFS to ZFS (this should start ringing alarm bells for many of you)....

As drives have been getting larger and larger, having a 512B block size becomes inefficient. Drive manufacturers have slowly been moving to a 4KB block size, also known as Advanced format. An interesting article on the transition can be found over at Anadtech: “Western Digital’s Advanced Format: The 4K Sector Transition Begins” You can determine if your drive using 4KB sector sizes by using smartctl: 1 2 # smartctl --all /dev/sdb | grep "Sector Size" Sector Sizes: 512 bytes logical, 4096 bytes physical When creating a zpool using ZFS, it queries the drive for the physical sector size to attempt to optimize disk operation....

I am the proud owner of a Quantum Superload 3, with a HP LTO-4 drive. The host connectivity is by SCSI, which I’ve never really had to use before. The following is the tale of me getting it to work. SCSI Controller The SCSI Controller I have is an QLogic card and shows up in lspci as follows: 1 2 07:04.0 SCSI storage controller: QLogic Corp. ISP10160 Single Channel Ultra3 SCSI Processor (rev 06) 07:05....

It’s well known that IPS has replaced SYSV in Solaris 11. It is still possible to install SYSV style packages. But how do these 2 package managers work together. If I use the netcat package as an example, I can see it as a IPS package and a SYSV package: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 df3804@empire:~$ pkg info netcat Name: network/netcat Summary: Netcat command Description: The nc(1) or netcat(1) utility can open TCP connections, send UDP packets, listen on arbitrary TCP and UDP ports and perform port scanning....

Configuring Solaris as a Kerberos client to work against a MS Active Directory Domain is documented by Oracle in the How to Configure a Kerberos Client for an Active Directory Server. This uses the /usr/sbin/kclient script provided by the kerberos-5 package, but makes some assumptions: The computer object will be created in OU=Computers The user joining the client to the domain must have Domain Administrator rights The first issue is fairly trivial to work around, but the second is more interesting and that is what the rest of this post will focus on....

The Real World Configuration Management was my first official event at LISA ‘12, with ~40 people sat around the table discussing their experiences and problems with configuration management. Experience was wide ranging in the room, from someone who was still evaluating configuration management tools, to Authors of CFEngine, LCFG & bcfg2. The day had a loose structure, broken down in to 3 parts. Discussion often went off-track, but that was to be expected....

Oracle provide a wealth of information for using various Naming and Directory Services with a Solaris OS as the client. Sometimes it’s hard to see the wood for the trees, therefore the following attempts to explain how to configure your Solaris 11 client to get user (passwd) and group (group) information from an OpenLDAP server. I assume you don’t have full control over your OpenLDAP server, or don’t want to bend over backwards to accommodate the Solaris client....

I woke up this morning to find 2 disks in my RAID5 array had been marked as failed! 1 2 3 4 5 6 7 8 9 10 11 12 13 md1 : active raid5 sda1[0] sdd1[4](F) sdc1[2](F) sdb1[1] 2930276352 blocks super 1.2 level 5, 512k chunk, algorithm 2 [4/2] [UU__] Jun 12 01:00:59 cube kernel: [12609415.780056] ata4: lost interrupt (Status 0x50) Jun 12 01:00:59 cube kernel: [12609415.843792] end_request: I/O error, dev sdd, sector 71 Jun 12 01:00:59 cube kernel: [12609415....