Re: [SLUG] Adaptec RAID problems
Thanks for your persistence Ben, I reduced the size of the array to a 2Gb RAID 1 and experienced the same problem. I've worked out that I can load more recent firmware using a more recent version of Sun Common Array Manager. The firmware load failed in RHEL so I'll have to wait until Sun has sorted that out. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Adaptec RAID problems
Hi Nick, Actually, I originally had 5x1Tb in a LVM group and that suited me fine. The HBA didn't support RAID for some reason. The only problem was that Fedora 11 udev tended to timeout when attempting to detect the individual disks. The only way to solve the problem was to plug in the SAS cable after the Fedora had booted and the detection was successful. Here is a post on fedora forum and entry in bugzilla: http://forums.fedoraforum.org/showthread.php?t=243493highlight=J4200 https://bugzilla.redhat.com/show_bug.cgi?id=582094 We upgraded the HBA to one which supported RAID and I decided on Fedora 13 (since Fedora 11 is now EOL). In Fedora 13, udev doesn't appear to detect individual disks. I'm guessing that the default behaviour for the new Adaptec driver in this kernel is not to expose physical components of the arrays (expose_physicals=0). I can probably try to override the driver preference and see what the IO is like to individual disks. Perhaps the problem will go away once I've installed the necessary firmware. I'm very tempted to run a chroot Gentoo from within RedHat. I really need a recent version of linux for development purposes, but it needs to sit on a vendor supported OS. I brought up a chroot image as a way to distribute GEANT -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Adaptec RAID problems
Hi, Thanks for the pointers. Is it possible that the HBA driver loads the firmware onto the drives? I've tried the harddrive vendor (Hitachi), but they don't release firmware updates. I reduced the size of the array to 4x2Tb RAID10 (4Tb redundant). Perhaps I'll try a smaller size to see if I have the same problem. The easiest way to replicate the problem is to run hdparm -t /dev/sda I get about 400kB/s on the problem RAID. The aacraid driver seems to mask single disks. AFAIK This prevents the driver timingout during bootup I've removed all LVMs so I can use modprobe and rmmod to experiment with different module options without restarting. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] Adaptec RAID problems
Hi, I am attempting to run Fedora 13 on a SunFire X4140 with an external Sun StorageTek enclosure with 7x2Tb disks. Fedora 13 boots of a single internal SAS drive. The HBA is an Adaptec AAC-RAID (rev 09). I have successfully created a RAID array using the adaptec BIOS. I have also successfully created a logical volume which is 7.99Tb. I attempted to create a filsystem on the the 7x2Tb RAID array from Fedora 13 but it didn't work. The computer hangs periodically and generates kernel warnings. I suspected that it is an issue related to the Adaptec RAID driver, so I downloaded the supported drivers from Sun. http://www.intel.com/support/go/sunraid.htm The most recent drivers for the external HBA adapter as supplied by Sun are at the level aacraid_1.1.5_2463. I compiled these into the Fedora kernel today and recreated the filesystem, however, I had the same problem. I have installed RedHat EL5 on another partition to help with diagnosing the problem. I've rebooted into Redhat and created the filesystem using the same command. This seemed to run perfectly fine (however I stopped it early because it would take several hours to complete). The baseline version of the Adaptec driver installed by Redhat EL5 is Adaptec aacraid driver 1.1-5[2437]-mh4 It appears that there are no hardware issues with the 7x2Tb array. I would prefer not to upgrade the firmware because it seemed to cause problems with RedHat EL5. Is there anyone in Sydney who had experience with running Fedora on Sun/Oracle Hardware? Thanks Robbie. Here are the Fedora 13 errors below. Code: [r...@petfire]# mkfs.ext3 -L mars /dev/mapper/vg_mars-lv_mars mke2fs 1.41.10 (10-Feb-2009) Filesystem label=mars OS type: Linux Block size=4096 (log=2) Fragment size=4096 (log=2) Stride=0 blocks, Stripe width=0 blocks 536207360 inodes, 2144799744 blocks 107239987 blocks (5.00%) reserved for the super user First data block=0 Maximum filesystem blocks=4294967296 65455 block groups 32768 blocks per group, 32768 fragments per group 8192 inodes per group Superblock backups stored on blocks: 32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208, 4096000, 7962624, 11239424, 2048, 23887872, 71663616, 78675968, 10240, 214990848, 51200, 550731776, 644972544, 1934917632 Writing inode tables: 811/65455 Message from sysl...@localhost at Jul 13 12:15:46 ... kernel:Stack: Message from sysl...@localhost at Jul 13 12:15:46 ... kernel:Call Trace: Message from sysl...@localhost at Jul 13 12:15:46 ... kernel: IRQ Message from sysl...@localhost at Jul 13 12:15:46 ... kernel: EOI Message from sysl...@localhost at Jul 13 12:15:46 ... kernel:Code: e8 91 ed ff ff 85 c0 74 17 49 8b 84 24 60 01 00 00 89 58 44 49 8b 84 24 60 01 00 00 44 89 68 2c 49 8b 84 24 60 01 00 00 8b 58 44 83 fb ff 75 cb b8 01 00 00 00 5b 5b 41 5c 41 5d c9 c3 55 48 89 817/65455 -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] IPv6 question
Just for fun, I'm setting up a FC10 machine with as an IPv6 router. The tunnel device (tun0) is automatically created using a script (aiccu) I want to allow other machines to use this box as a gateway, but I think I have the config wrong 1: lo: LOOPBACK,UP,LOWER_UP mtu 16436 inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eth0: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qlen 100 inet6 2001:1938:83:52::3/64 scope global valid_lft forever preferred_lft forever inet6 fe80::20d:57ff:fd9b:4b24/64 scope link valid_lft forever preferred_lft forever 40: tun0: POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP mtu 1280 qlen 500 inet6 2001:1938:83:52::2/64 scope global valid_lft forever preferred_lft forever inet6 fe80::1838:83:52:2/64 scope link valid_lft forever preferred_lft forever 2001:1938:83:52::/64 dev eth0 proto kernel metric 256 mtu 1500 advmss 1440 hoplimit 4294967295 2001:1938:83:52::/64 dev tun0 proto kernel metric 256 mtu 1280 advmss 1220 hoplimit 4294967295 fe80::/64 dev eth0 proto kernel metric 256 mtu 1500 advmss 1440 hoplimit 4294967295 fe80::/64 dev tun0 proto kernel metric 256 mtu 1280 advmss 1220 hoplimit 4294967295 default via 2001:1938:83:52::1 dev tun0 metric 1024 mtu 1280 advmss 1220 hoplimit 4294967295 I have radvd running and everything works well on the LAN using the stateless addresses. -- I've tried many combinations of a few things to solve the problem turning off /etc/init.d/iptables turning off /etc/init.d/ip6tables sysctl -w net.ipv6.conf.default.forwarding=1 sysctl -w net.ipv6.conf.all.forwarding=1 Restricting prefix of tun0 # /sbin/ip -6 addr del 2001:1938:83:52::2/64 dev tun0 # /sbin/ip -6 addr add 2001:1938:83:52::2/126 dev tun0 Changing the global ip on eth0 to the same as tun0 Any thoughts? Thanks Robbie Barnett rob...@barnett.id.au 0431864709 Skype Name: retsil42 -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] Fortress .... err Firewall Australia
I am really most concerned about all the side effects of filtering that we are yet to uncover. For example: If I hosted on a dynamic IP, although I have been assured that it is unlikely, how would I deal with the event that my IP has been blocked to other users? Could the federal government consider not blocking domestic addresses and actually enforce Australian law on our own turf? What if I ran a VPN on port 80? How do I even know that I'm on the blocked list? What about lesser known sites such as the Internet WayBack machine? Would any objectionable material result in a blanket ban? Robbie -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] Keeping wife on linux
My wife and I have a shared computer at home, however, I seem to be facing a loosing battle for her experiences with linux to remain pleasant. I am running fedora core 9 with additional packages from the Livna repository Some nuances: * We've not been able to find a way to purchase songs from Yahoo or iTunes. I've tried foxy tunes and wine without much success. * We've had some difficulty with video codecs. Channel nine or ten is using a codec which includes advertisements but only works for Windows Media Player 10+. * I've also played a few DVDs in MPlayer and other players and just get the "Downloading movies is stealing" on continuous repeat. Is there a distro which is more likely to work with proprietary codecs? Or am I just better to stick to FC or possibly Ubuntu and use 3rd party packages as i have been doing? I've been toying with using vmware and running XP SP2, but I think that it would leave me with two systems to maintain rather than one. I may also have to buy a new machine (dual core) to meet the system requirements. Any suggestions? -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] Safe samba configuration
I use samba on a mixed OS network. We have several NAS devices and two Fedora workstations which share drives using NAS. I have configured the Fedora workstations to only announce on the local subnet (/24) to prevent inadvertent connections from the wider organization. I implement this by putting the following in smb.conf remote announce = 10.9.136.0/24 I also set local master = no Because I figured that there were other computers more suited to being the master browser (whatever that means) Unfortunately I've been faced with some windows boxes spontaneously not able to browse or connect to the network. It is occasional, but seemed to decrease in occurance when I reduced the use of SMB shares. I'm really keen to find a safe configuration for my SMB shares, but I'm not sure where I would find how to do this. It's very difficult to diagnose which configuration works better without months of investigation. Any help would be greatly appreciated. -- Robbie Barnett [EMAIL PROTECTED] 0431864709 Skype Name: retsil42 -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] SLUG January meeting video (Microsoft QA) is online
Thanks for putting the January meeting online on so promptly. For me, the material has sparked a few interesting conversations with friends in the industry. I've noticed how MS folk tout their openness as though they have solved the interoperability problem in one fowl swoop. I wouldn't want to cause offence to the great people who work there, but in reality, they have only offered a very small window into Microsoft technology under the Open Specification Promise (OSP). Other companies such as Sun Microsystems have been put into a position that they must make some of their largest projects open source in order to have a hope of market share. We live in very interesting times indeed... Thanks, Robbie -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] HFS+ support
I suspect that the cause is poor support for journalling. I created a separate HFS+ partition with journalling turned off. This time I didn't have to force rw mode, so I suspect it will be more stable. I'll post back here if it dies again. Thanks Robbie -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] HFS+ support
Hello, I've recently set up a Mac/Linux dual boot system. I mounted the HFS+ drive using the experimental HFS driver in read/write mode so I could share my user home between both operating systems. Unfortunately this corrupted my HFS file system today and I had lots of lively messages coming from the Linux kernel. I was admittedly in a bit of a rush to get it going, but now I have the time to relfect on my mistakes. Would anyone recommend a way of running Mac/Linux dual boot? How would you share userspace files? Thanks Robbie -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] RedHat at UOW
Hi, I have a colleague from Uppsala University, Sweden. He uses ITK/VTK for data processing and manipulation of Medical Images. He currently has a post-doc position which is shared between UOW and a Sydney hospital. He's informed me that UOW would not let him use his RedHat (Fedora Core 3) machine on the campus network because they only allow for approved OSes to connect to the network. Does anyone have any ideas about how to appeal this decision/policy? I completed a postgrad course at UOW. In all my time there I never managed to get any response from the IT department with regards to policy. Regards, Robbie -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
