From warren.liang at cox.net Thu May 5 19:24:35 2011 From: warren.liang at cox.net (Warren Liang) Date: Thu, 5 May 2011 16:24:35 -0700 Subject: SUMMARY: Cannot startup samba Message-ID: <20110505192435.Y161I.393102.imail@fed1rmwml44> Hello: Thanks to Ben Cottle. He provides the solution: This is because the smb.conf file was moved by Oracle from /etc/sfw/smb.conf to /etc/samba/smb.conf Just copy/move your conf file across and clear the maintenance on your services # svcadm clear svc:/network/samba:default This one got me as well .. Regards, Ben Cottle -----Original Message----- After install the latest 10_Recommend patch (144488-12. release date: Apr 21, 2011), the following services fail to start: May 5 15:14:00 svc.startd[7]: svc:/network/wins:default: Method "/usr/sbin/nmbd -D" failed with exit status 1. May 5 15:14:00 svc.startd[7]: network/wins:default failed: transitioned to maintenance (see 'svcs -xv' for details) May 5 15:14:00 svc.startd[7]: svc:/network/samba:default: Method "/usr/sbin/smbd -D" failed with exit status 1. May 5 15:14:01 svc.startd[7]: network/samba:default failed: transitioned to maintenance (see 'svcs -xv' for details) _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From sanelson at gmail.com Sat May 7 06:15:13 2011 From: sanelson at gmail.com (Stephen Nelson-Smith) Date: Sat, 7 May 2011 11:15:13 +0100 Subject: SUMMARY: Unable to reboot after LU Message-ID: Hello managers, Thanks for everyone's suggestions. In the end I couldn't get the machine to reboot sanely at all, and had to power cycle, after which it came up cleaning into the new BE anyway. However some of your advice didn't come in time. Herewith a summary of lessons learned: * Ensure that after lucreate and luupgrade, that luactivate has been run. I did actually do this. * Cal Dunigan points out that it's probably good practice not to do a big jump from 6 to 9, and take intermediate steps. I'd already done one machine this morning straight from 6 to 9 and it worked fine, but I think this is good advice. * Be patient - often a reboot can take many minutes to take effect. I've experienced this too and had waited more than 30 mins before starting to investigate. * Read the release notes! Had I done so (as John Leadeham did) I would have found that a hang is known to occur while svc.startd halts services. I saw exactly the evidence that the notes point out, so I suspect this was the problem. For completeness, the workaround is : # /usr/share/webconsole/private/bin/wcremove -i console # svcadm clear system/webconsole:console # smcwebserver start Here's the original problem statement: ----- I've just completed a live upgrade (on ZFS root) from 10u6 to 10u9: # lustatus Boot Environment Is Active Active Can Copy Name Complete Now On Reboot Delete Status -------------------------- -------- ------ --------- ------ ---------- sol10u6 yes yes no no - sol10u9 yes no yes no - However, despite shutting down the oracle databases on this machine, plus the other application, neither init 6 nor shutdown -y -i6 -g0 seems to have any effect. who -r shows run level 6: . run-level 6 May 6 23:27 6 0 3 Unless I'm missing something obvious (which is possible - it's late), the process table doesn't appear to show any SMF service hanging: UID PID PPID C STIME TTY TIME CMD root 0 0 0 Oct 26 ? 0:41 sched root 1 0 0 Oct 26 ? 26:48 /sbin/init root 2 0 0 Oct 26 ? 0:01 pageout root 3 0 0 Oct 26 ? 6647:28 fsflush root 9 1 0 Oct 26 ? 27:58 /lib/svc/bin/svc.configd daemon 138 1 0 Oct 26 ? 1828:54 /usr/lib/crypto/kcfd root 78 1 0 Oct 26 ? 0:00 devfsadmd stephen 15304 15303 0 10:58:29 ? 1:26 /usr/lib/ssh/sshd daemon 340 1 0 Oct 26 ? 0:00 /usr/sbin/rpcbind root 137 1 0 Oct 26 ? 0:00 /usr/lib/sysevent/syseventd root 24264 1 0 23:28:15 ? 0:03 /lib/svc/bin/svc.startd root 377 1 0 Oct 26 ? 2:53 /usr/sbin/syslogd root 24644 10661 0 00:17:31 pts/1 0:00 ps -ef root 24413 24264 0 23:42:17 console 0:00 -bash root 24261 1 0 23:28:13 ? 0:00 /usr/lib/power/powerd oracle 1212 1 0 Oct 26 ? 40:45 /u01/app/oracle/product/10.2.0/db_1/bin/tnslsnr LISTENER -inherit root 10661 15310 0 20:10:10 pts/1 0:01 -bash root 19317 1 0 Nov 20 ? 1075:33 /usr/java/bin/java -Djava.util.logging.config.file=/var/apache/tomcat55/conf/lo root 19324 1 0 Nov 20 ? 35:56 /u01/app/uniface/usys8404/bin/urouter stephen 15310 15304 0 10:58:32 pts/1 0:00 -bash root 15303 1 0 10:58:29 ? 0:00 /usr/lib/ssh/sshd I'm a bit stuck. I'm concerned that if the machine is restarted without the proper init process, the new BE won't boot, and I'm not sure what will happen. ----- Thanks again for everyone's help. S. _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From Rob.McMahon at warwick.ac.uk Tue May 17 04:59:01 2011 From: Rob.McMahon at warwick.ac.uk (Rob McMahon) Date: Tue, 17 May 2011 09:59:01 +0100 Subject: SUMMARY: Using raidctl to mirror boot disk after installation. In-Reply-To: <4DD14834.1000504@warwick.ac.uk> References: <20110510043002.0426E361E0@apps0.cs.toronto.edu> <4DD14834.1000504@warwick.ac.uk> Message-ID: <4DD238D5.5000203@warwick.ac.uk> On 16/05/2011 16:52, Rob McMahon wrote: > We have a Sun Fire X4100 M2, which has just lost a disk. This was all > mirrored using SVM, but getting the system back running on one disk > proved to be more of a pain than it should have been, and we'd like to > use the hardware RAID controller when we replace the broken disk. The > system is now using normal raw partitions as a result of the problems > bringing it back. Now I'd like to think I could just do > > raidctl -c c3t2d0 c3t3d0 > > and be done, but there are dire warnings that you have to 1) do this > from remote / DVD media (which is not a problem), and 2) re-install the > OS afterwards, which is. > > It's a long time since I've done this, but I remember the disk geometry > changes. If we just lose two or three cylinders off the end of the > disk, this wouldn't be a problem, since the last 3 cylinders are just a > metadb, which I could destroy. > > Has anyone got any advice ? This is quite urgent, because if I can't > find an answer quickly, I'll just have to SVM it up again to be safe. > > Cheers, > > Rob > Almost unanimously people have said stick with SVM. The glaring error I made, which I realised myself as I was digging the machine out of its hole, was to miss out adding the following to /etc/system set md:mirrored_root_flag=1 Matt Morris pointed this out. Thanks. One of the partitions (with scratch data - quarantined virus mails) was created as RAID 0 for performance rather than RAID 1 for redundancy, and this was obviously an issue too. We live and learn. Excerpts from the responses (thanks everyone): Chris Hoogendyk, Adrian Koester, and Matthew Stier all basically said you do have to re-partition and re-install. Anthony D'Atri said forget HW RAID and go ZFS. It's a good point, but these machines were installed while ZFS was still young, and there was no ZFS boot. Another re-install. In addition to the /etc/system addition, Matt Morris said "And hardware RAID is AWFUL, IMO. SVM is tried and true, easy to manage, and easy to get drive status online. I've just never had good experiences with HW RAID..." A sentiment I agree with, but people around me were asking why I didn't go for HW RAID. I'm sticking with SVM. Thanks again for all your input. Rob -- E-Mail: Rob.McMahon at warwick.ac.uk PHONE: +44 24 7652 3037 Rob McMahon, IT Services, Warwick University, Coventry, CV4 7AL, England _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From rob.de.langhe at twistfare.be Wed May 18 16:46:16 2011 From: rob.de.langhe at twistfare.be (Rob De Langhe) Date: Wed, 18 May 2011 22:46:16 +0200 Subject: SUMMARY: (very) large disks in Netra-X1 ? In-Reply-To: <20110518224201.31913o7elc3iy2io@www.twistfare.be> References: <20110518224201.31913o7elc3iy2io@www.twistfare.be> Message-ID: <20110518224616.16674yw9vs276i4o@www.twistfare.be> got it myself already : http://supportforum.sun.com/hardware/index.php?t=msg&goto=7912&rid=0#msg_7912 reminds me that the 24-bit IDE controller can address only the first 137 GB of a disk too bad... Citeren Rob De Langhe : > hi all, > > just wondering if someone knows where to find the maximum internal > disk capacity that the PROM of a Netra-X1 can handle ? > I currently have 2 disks installed, but one of them is 200 GB Western > Digital and it shows up as 127 GB in "format"... > > I got already while ago the confirmation via this wonderful forum > that a Netra-X1 can only recognize the first 128 GB or so max > capacity of a disk, and no more. > > Is there by any magical way a method found by someone to increase > that limit, so that larger disks are available with their full > capacity to a Netra-X1 ? > > silently wishful thinking... > > Rob > > > ---------------------------------------------------------------- > This message was sent using IMP, the Internet Messaging Program. > _______________________________________________ > sunmanagers mailing list > sunmanagers at sunmanagers.org > http://www.sunmanagers.org/mailman/listinfo/sunmanagers > ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From rob.de.langhe at twistfare.be Thu May 19 04:19:45 2011 From: rob.de.langhe at twistfare.be (Rob De Langhe) Date: Thu, 19 May 2011 10:19:45 +0200 Subject: SUMMARY: international PSU in SUN X2100 servers ? In-Reply-To: <20110518230427.607717iuiu04dvfk@www.twistfare.be> References: <20110518230427.607717iuiu04dvfk@www.twistfare.be> Message-ID: <20110519101945.75876m9vzpmylako@www.twistfare.be> I really love this forum... Within the hour, 12 people confirmed that the PSUs shipped in SUN servers have since long been self-adjusting for the input Voltage they get . So just need a localized power cord, and can hop around with the same server across the world, from USA over Europe to Asia, spend some lazy time on the Aussie beach, and play tunes on it in Africa. What a shame that SUN management made the company broke so that this wonderful technology got sold out to Oracle... thx all ! Rob Citeren Rob De Langhe : > hi all, > > just wondering if there is a difference in the power supply installed > in SUN-X2100 servers shipped in the USA compared to when shipped in > Europe ? > I read from the Hardware Specifications on this (EOL, I know) server > that its needs 90-264 V AC (47-63 Hz) as AC power. > Does that mean that a model shipped in the USA (110 Volts, if it > hasn't changed meanwhile) can be equally used in Europe (220-250 > Volts), provided local power cable (duh) ? > > Rob > > > ---------------------------------------------------------------- > This message was sent using IMP, the Internet Messaging Program. > _______________________________________________ > sunmanagers mailing list > sunmanagers at sunmanagers.org > http://www.sunmanagers.org/mailman/listinfo/sunmanagers > ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From Mark.Twardzik at jhuapl.edu Thu May 19 08:25:04 2011 From: Mark.Twardzik at jhuapl.edu (Twardzik, Mark J.) Date: Thu, 19 May 2011 08:25:04 -0400 Subject: SUMMARY: Authenticating Solaris 10 through Active Directory In-Reply-To: References: Message-ID: Thanks everyone for all the replies. I tried using Likewise Open as suggested by Brandon below, and it worked for my situation. It was much simpler and faster than trying to manually configure everything. Very good documentation as well. Thanks again! Mark _____________________________________________________________________________ _________________________________________________________________________ Hi Mark, What we've recently started working with is the Likewise Open Source AD package for Solaris and have had a good amount of success. We haven't done a full rollout yet as we still need to do more testing but this could be of some use to you. The page is at: http://www.likewise.com/products/likewise_open/index.php Hope that helps, -Brandon -- Brandon Battis _____________________________________________________________________________ ___________________________________________________________________ Thanks for the quick reply David. I have already been to those websites and used some of the information. I believe it's just a matter of properly configuring the ldap_client_file. Haven't tried it myself, but you may want to check out: http://blog.scottlowe.org/2007/04/25/solaris-10-ad-integration-version-3/ http://blog.scottlowe.org/2008/11/19/no-solaris-ad-integration-update/ It's a bit old now, but the principles should still apply. David Magda _____________________________________________________________________________ _____________________________________________________________________ are you sure that the "proxy" account which you use, is defined in Active Directory to allow a "simple" password authentication ? If I recall, per default it will expect the Windows-style authentication, so will fail Rob, Thanks for the quick reply. My ldap_client_file has a few different attributes, such as NS_LDAP_AUTH= sasl/GSSAPI and NS_LDAP_CREDENTIAL_LEVEL= self. I had previously tried using simple and proxy, but that didn't work either. Snoop output a different error message with proxy and simple: *[LDAPMessage] > [message ID] > Operation *[APPL 1: Bind Response] > [Result Code] > 1 Invalid Credentials > [Matched DN] > [Error Message] > 80090308: LdapErr: DSID-0C0903A9 , comment: AcceptSecurityContext I'm hoping it might just be a matter of tweaking the ldap_client_file. Thanks again, Mark Mark, you need a "bind" (or "proxy") account (and its password) in the LDAP (Active Directory) repository, that your Solaris clients can use to request data from the AD This "bind" account is what you will use in the command to manually configure a Solaris-10 client as LDAP client: ldapclient -v manual -a defaultServerList=kdc.ourdomain.internal -a defaultSearchBase="dc=ourdomain,dc=internal" -a authenticationMethod=simple -a followReferrals=FALSE -a defaultSearchScope=one \ -a searchTimeLimit=30 -a credentialLevel=proxy \ -a proxyDN="cn=bindaccount,ou=Process,ou=Logins,ou=THISDEPT,dc=ourdomain,dc=inte rnal" \ -a proxyPassword=somepwd \ #-a objectclassMap=passwd:posixAccount=user \ -a serviceSearchDescriptor=passwd:ou=PER,OU=People,OU=THISDEPT,DC=OURDOMAIN,DC=I NTERNAL \ -a attributeMap=passwd:homeDirectory=unixHomeDirectory This should result in a file "/var/ldap/ldap_client_file" with the following contents : NS_LDAP_FILE_VERSION= 2.0 NS_LDAP_SERVERS= kdc.ourdomain.internal NS_LDAP_SEARCH_BASEDN= dc=ourdomain,dc=internal NS_LDAP_AUTH= simple NS_LDAP_SEARCH_REF= FALSE NS_LDAP_SEARCH_SCOPE= one NS_LDAP_SEARCH_TIME= 30 NS_LDAP_CACHETTL= 0 NS_LDAP_CREDENTIAL_LEVEL= proxy NS_LDAP_SERVICE_SEARCH_DESC= passwd:ou=PER,OU=People,OU=THISDEPT,DC=OURDOMAIN,DC=INTERNAL NS_LDAP_OBJECTCLASSMAP= passwd:posixAccount=user There are some additional important steps to take (like editing /etc/pam.conf and /etc/nsswitch.conf), but I guess you figured these already out. If not, I can inform you of more. good luck Rob Rob De Langhe -----Original Message----- From: sunmanagers-bounces at sunmanagers.org [mailto:sunmanagers-bounces at sunmanagers.org] On Behalf Of Twardzik, Mark J. Sent: Thursday, May 12, 2011 3:57 PM To: sunmanagers at sunmanagers.org Subject: Authenticating Solaris 10 through Active Directory I have a Sun Netra T5440 SPARC running Solaris 5.10 with a fresh End User installation, no additional patches. I am trying to authenticate it through Active Directory on a Windows Server 2008 R2 system. I followed the instructions in Sun document 'Using Kerberos to Authenticate a SolarisTM 10 OS LDAP Client With Microsoft Active Directory' , although I had to change a few things to successfully navigate some of the steps. This was expected, as I realize the document was written for Server 2003. I believe Kerberos is configured properly, as 'kinit (test user)' obtains tickets according to klist. However, neither 'ldaplist -l passwd (test user)' nor 'getent passwd (test user)' work. It looks like a binding issue with sasl/GSSAPI based on the following error messages: 'cat /var/adm/messages | grep ldap' results contain the repeated error 'libsldap: makeConnection: failed to open connection using sasl/GSSAPI to ForestDnsZones.my.domain' 'snoop -v | grep -I ldap' contains *[LDAPMessage] [message ID] Operation *[APPL 1: Bind Response] [Result Code] Success [Matched DN] [Error Message] SASL Credentials [7] Any help would be greatly appreciated, as I have spent a great deal of time looking through message boards and playing with different configurations. Send email and I will summarize to the list. _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From scooper at hp.com Thu May 19 10:08:48 2011 From: scooper at hp.com (Cooper, Scott) Date: Thu, 19 May 2011 14:08:48 +0000 Subject: SUMMARY: Dhcp server in Solaris 8 container In-Reply-To: <7F8C20DA3E7E2844990DAE9E00285B4A4B409F383C@GVW1363EXC.americas.hpqcorp.net> References: <7F8C20DA3E7E2844990DAE9E00285B4A4B409F383C@GVW1363EXC.americas.hpqcorp.net> Message-ID: <7F8C20DA3E7E2844990DAE9E00285B4A4B409F3B5B@GVW1363EXC.americas.hpqcorp.net> Al, The answer is to make the port available to the non global zone exclusively. Set ip-type=exclusive. Dhcp on the non global zone then works just fine. Thanks to: Rob De Langhe & Matthew Stier for pointing me in the right direction. Regards Scott -----Original Message----- From: sunmanagers-bounces at sunmanagers.org [mailto:sunmanagers-bounces at sunmanagers.org] On Behalf Of Cooper, Scott Sent: 19 May 2011 09:31 To: sunmanagers at sunmanagers.org Subject: Dhcp server in Solaris 8 container Hi all, We have a legacy system running Solaris 8 that we have ported into a solaris 8 container. The server has always been a dhcp server however within the zone it fails to start with a message - Error discovering number of network interfaces: no such device or address. Is what I am trying to do supported? Or will I need to try and move the dhcp serving to the Global zone? Regards Scott Cooper Technical Consultant HP Enterprise Services Defence & Security UK Ltd. Email: scooper at hp.com Hook D, Bartley Wood Business Park, Hook, Hampshire, RG27 9XA This email contains information which is confidential and may be privileged. Unless you are the intended addressee (or authorised to receive for the addressee) you may not use, forward, copy or disclose to anyone this email or any information contained in this email. If you have received this email in error, please advise the sender by reply email immediately and delete this email. HP Enterprise Services Defence & Security UK Ltd Registered Office: Cain Road, Bracknell, Berkshire, RG12 1HN. Registered in England No: 936751 _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From WHITERL at nv.doe.gov Fri May 20 11:56:39 2011 From: WHITERL at nv.doe.gov (White, Bob (CONTR)) Date: Fri, 20 May 2011 15:56:39 +0000 Subject: SUMMARY: Configure sendmail for Exchange 2008 - MORE INFO In-Reply-To: Message-ID: I apologize for the delay in sending this. Been hiding out in the Nevada desert on field work. Anyway, the problem has been fixed. All of the tests that people suggested showed that both the Sun and the Exchange server were configured correctly, and something was accepting all the e-mails. Sal Serafino suggested a verbose debug logging of the sendmail to see exactly what is going on. At first it looked like everything was indeed being accepted by the mailhub. Then I noticed a small error. Sendmail was attempting to relay to mailhost.my.domain. That address wasn't being found in the DNS. When I added an entry to the hosts file for mailhost, using the Exchange server address, all the mail came cascading in. Not sure why that works, or how sendmail was working before with only mailhub defined, but there you have it. One small change. Thanks to everyone for your help and suggestions. On 4/22/11 7:21 AM, "White, Bob (CONTR)" wrote: >Up until last month we were still running Exchange 2003. My two Solaris >10 >boxes were set up to send only to a mailhub. All was working fine. Then >we >went to 2008. Now the mail no longer works. Due to some other issues, >the >network and IT people are real busy putting out some major fires, and >don't >have time to look at my minor issue. > >Is there some special option that needs to be enabled to get sendmail to >send >to a 2008 server? > >Solaris 10 fully patched on a Sun Ultra 20. > >No longer works = messages go out, and disappear. No error messages from >sendmail. The mail just never arrives at destination address. > >Thanks > >Bob White >M/S NLV075 >National Security Technologies, LLC >Contractor for the United States Department of Energy >Office: 702-295-2939 >Cell: 702-630-0352 >Pager: 702-794-1809 >Fax: 702-295-2934 >whiterl at nv.doe.gov >_______________________________________________ >sunmanagers mailing list >sunmanagers at sunmanagers.org >http://www.sunmanagers.org/mailman/listinfo/sunmanagers _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From aad at dreamsnake.net Mon May 23 05:43:11 2011 From: aad at dreamsnake.net (Anthony D'Atri) Date: Mon, 23 May 2011 02:43:11 -0700 Subject: SUMMARY: JBOD / passthrough mode on LSI MegaRAID SAS 9261-8 aka Sun Storage 6 Gb/s SAS PCIe RAID HBA SG-SAS6-R-INT-z ? In-Reply-To: References: Message-ID: Preface: why on earth does Sun / Oracle default to goofy RAID HBA's when ZFS and even SVM provide superior mirroring? AFAICT there isn't a functional way to interrogate the status of RAID volumes on this HBA from the OS -- how is one supposed to monitor for disk failures? I ended up managing to delete whatever volume came from the factory via the misnamed WebBIOS. Fired up the Java remote console and fished around via and Enter until I found an option to do so. The installation guide for this card only touches on the bizarre pre-boot CLI, but a statement that the syntax is the same as the MegaCLI utility led me to a user's guide for the latter, from which I synthesized that -cfgldadd -r0[252:0] -a0 -cfgldadd -r0[252:1] -a0 -cfgldadd -r0[252:2] -a0 -cfgldadd -r0[252:3] -a0 performs the workaround of creating four degenerate whole-disk stripes that can be seen by BIOS and the OS, and thence ZFS'd. There is of course the usual uncertainty about booting off the mirror if the primary fails, thanks to the miserable legacy of BIOS. A bit of clarification of the above syntax: on SPARC hosts, commands are prefixed with "cli", and one selects an HBA from potentially several ahead of time, much like one does with ndd on certain older ethernet interfaces. On x86 hosts every command has to specify the desired HBA. I guessed that the first would be 0, and that worked. The 252 seems to be the SES enclosure ID for internal disks; 0-4 should be obvious. The moral of the story is to be very, very sure that new systems you order come with normal, non-RAID HBA's. > I've searched the web and found only others asking this, with a few hints but no answers. > > I have an x4170m2 that was purchased by someone outside the company that came with one of these pointless RAID HBA's. I'm trying to figure out how to get the thing to simply present the four disks attached to it to BIOS and the OS as-is, without RAID games, so that I can set up customary ZFS goodness on them. > > I entered a ticket with Soracle, and so far the tech's response has been that he's never heard of this card -- despite the fact that it's what the system ships with by default! > > The BIOS configuration utility presented by the card is even worse than the typical PC-oriented card: the textual interface is a CLI that's cryptic beyond belief, and "WebBIOS" turns out to only work on a graphical console, with the usual limited compatibility with ILOM's remote console utility (eg. the mouse doesn't work). I managed to destroy whatever volume it had configured from the factory, but can't figure out how to get it to pass through the disks, or even set up stupid one-disk virtual volumes. > > I tried booting diskless and using megacli, but when I grab that from Intel's web site and run it, it doesn't appear to even see the HBA. The HBA at power-on does report four attached disks. > > So, I'm hoping that someone out there might have an idea of how to bypass the onboard RAID lunacy on this card so that I can get the system installed and in production. I'll summarize as appropriate. Thanks! _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From joe_fletcher at btconnect.com Sun May 29 05:59:55 2011 From: joe_fletcher at btconnect.com (joe fletcher) Date: Sun, 29 May 2011 10:59:55 +0100 Subject: SUMMARY: v490 vs M4000 References: <7FC614F7CE1FCE449C3674BAD9A52032AC7658@HEMV3BUKER.he.local> Message-ID: <7FC614F7CE1FCE449C3674BAD9A52032AC765C@HEMV3BUKER.he.local> Thanks to all the replies. There were many, especially peope offering me to sell me either the v490 memory or the servers. Ultimately the question has been answered by my senior management who have decided just to buy them. I still think I can get a better result with x86 but I won't be given the chance now. Factoring in the support costs etc does make it seem a little less painful, money-wise. We've got an M3000 under test now as a proof-of-concept. Concensus is that I should see a marked improvement, given the doubling of clock speeds, PCI-X buses etc. Allowing for "developer's sql" if you know what I mean, based on the replies I'm hoping for a 40-50% kick. I'll write up an addenda when we've got the testing completed. Again thanks for all the replies. Joe ________________________________ From: sunmanagers-bounces at sunmanagers.org on behalf of joe fletcher Sent: Fri 5/20/2011 5:22 PM To: sunmanagers at sunmanagers.org Subject: v490 vs M4000 Hi, Can anyone provide me with a real-world comparison of the old v490 vs the new M4000s? We've got a choice of upgrading 4 v490s to 64Gb RAM cost approx $100K vs buying some new M4000s, cost somewhere between $150-200K. Searches on benchmarks indicate performance improvements on the new architecture produce results indicating anywhere between 0 and 25%. Based purely on clock speeds we might expect 25% but as always, clock speed isn't that much of an indicator. Workload is multi-tiered mixture of java and a well known database though not one owned by Larry. If anyone has any pointers which suggest buy new or keep the old I'd appreciate them. At present, service contracts etc are outside the scope of the discussion. TIA Joe _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers