============================================================================== NCAR/CGD Systems Work Wed 30 Nov 05 6:00am - 8:00am Rack Arrangement ----------------- The rack re-arrangement in the CGD machine room continue: The following systems/services will be down during this time: CGDW2KCIT . Print services for Windows boxes ghotiol . IMAGe compute server routt . Failover e-mail server poorman . General use Linux systems intrepid . CMS compute server scp . External scp server cgdprint . Guest printing for Windows boxes haystack . CGD software repository navajo . /usr/local for goldhill goldhill . General use system, external access alef . Primary DNS and YP authentication server If this work will seriously impact any planned computing, please contact the CGD/IS staff at systems@cgd.ucar.edu so that we may re-schedule. -- ============================================================================== ============================================================================== NCAR/CGD Systems Work Wed 2 Nov 05 6:00am - 8:00am As always, please contact the CGD/IS staff at systems@cgd.ucar.edu if this work will seriously impact any planned computing. Rack Moves ---------- Musical machine room continues with more system moves. The following systems/services will be down during this period: Sanitas General Use Flagstaff General Use Salina General Use Bangkok Compute Cluster Calgary Compute Cluster Klondike /fs/cgd/home0 /fs/cgd/data0 /fs/cgd/csm /project/cas /project/cms /project/oce /project/cseg /project/bgc1 /project/bgc2 /project/satftp1 /project/satftp2 Paleo /project/paleo2 Gimel Primary DNS for some systems -- Mark --0- ---------------------------------------------------------------------- Mark Moore UCAR/NCAR/CGD mmoore@ucar.edu 1850 Table Mesa Drive (W) 303 497-1338 Boulder, CO 80305 (F) 303 497-1324 ============================================================================== ============================================================================== NCAR/CGD Systems Work Wed 26 Oct 05 6:00am - 8:00am Re-Racking ---------- The following systems will be down for rack cleanup: Bangkok HW Upgrade ---------- The following systems will be down for hardware upgrades: Flagstaff: 2 x 295 MHz CPU -> 2 x 450 MHz CPU 576 MB RAM -> 1024 MB RAM Sanitas: 2 x 295 MHz CPU -> 2 x 360 MHz CPU 768 MB RAM -> 1024 MB RAM OS Upgrade ---------- The following systems and services will be out (about 15 minutes) for and operating system upgrade. During this time, backup servers will handle the load, but will be somewhat slower (30 seconds). Beis: DNS (Domain Name Services) NIS (Network Information Systems) Gimel: DNS (Domain Name Services) NIS (Network Information Systems) -- Mark --0- ---------------------------------------------------------------------- Mark Moore UCAR/NCAR/CGD mmoore@ucar.edu 1850 Table Mesa Drive (W) 303 497-1338 Boulder, CO 80305 (F) 303 497-1324 ============================================================================== ============================================================================== NCAR/CGD SysWrk: Wed 19 Oct 05 Rack Moves ---------- CGD/IS is starting a massive rack re-organization in the machine room. This involves moving all systems out of a rack, rebuilding the rack, and then moving systems back into the rack. The improvements will be better cable management, installation of remote consoles, and installation of remote power management. The following servers (and services) will experience 10-20 min outages while they are relocated. Server Service ------ ------- Goldhill Outside access to CGD, general logins Wallstreet Weather forecasting contest Dalet DHCP service (laptops) Flatiron wreq Whiteboard CGD whiteboard viewing Mailhost E-mail, Webmail, IMAP(/S), POP(/S) Gimel DNS Beis DNS Cheyenne CGD performance and security monitoring Echorock Web, mailman services If this work will seriously impact any planned computing, please notify CGD/IS before COB Tue 18 Oct 05. -- Nick Wehrheim Systems Administrator UCAR/NCAR/CGD nickw@ucar.edu 1850 Table Mesa Drive (W) 303-497-1374 Boulder, CO 80305 (F) 303-497-1324 ============================================================================== ============================================================================== NCAR/CGD Systems Work 6:00am - 8:00am Wed 12 Oct 05 Rack Moves ---------- CGD/IS is starting a massive rack re-organization in the machine room. This involves moving all systems out of a rack, rebuilding the rack, and then moving systems back into the rack. The improvements will be better cable management, installation of remote consoles, and installation of remote power management. The following servers (and services) will experience short 5-10 min outages while they are relocated. Server Service ------ ------- Goldhill Outside access to CGD, general logins Wallstreet Weather forecasting contest Dalet DHCP service (laptops) Flatiron wreq Whiteboard CGD whiteboard viewing Mailhost E-mail, Webmail, IMAP(/S), POP(/S) Gimel DNS Beis DNS Cheyenne CGD performance and security monitoring Echorock Web, mailman services This is incredibly short notice. My apologies. If this work will seriously impact any planned computing, please notify CGD/IS before COB Tue 11 Oct 05. -- Mark --0- ---------------------------------------------------------------------- Mark Moore UCAR/NCAR/CGD mmoore@ucar.edu 1850 Table Mesa Drive (W) 303 497-1338 Boulder, CO 80305 (F) 303 497-1324 ============================================================================== ============================================================================== All CGD services will remain operational during the Oct. 8, 2005 powerdown. See below from SCD services that are affected. ============================================================= MESA LAB- SATURDAY, OCTOBER 8, 2005 FALL POWER DOWN ANNOUNCEMENT The Mesa Lab semi-annual fall power down is scheduled for October 8th. The computer room itself will be down from 0600 hrs to 1800 hrs, significant computing systems will be taken off line beginning at 0001 hrs – 10/8/2005. SCD will be performing a number of tasks along with Physical Plant Services during this outage. These tasks will have differing impacts to the computer room throughout the day but the major impact will be from 0630 to 0830 hrs. Beginning at 0630 hrs, SCD will test the EMERGENCY POWER OFF (EPO) system for the Mesa Lab computer room. This test will result in a TOTAL outage of the Mesa Lab computer room. During this time the following services will be unavailable: ALL network services at the Mesa Lab and Marshall including phones Services affected for ALL UCAR sites VPN RAS Wireless Voicemail Long Distance (outbound) Email Following the system test, all the breakers will be reset and some core systems will become available again such as phone and critical computing systems. The EPO test portion of the power down is scheduled to be completed and everything reset by 0830 hrs. VPN and FTP will not be available until after 1000 hrs. For the remainder of the day numerous tasks will be undertaken on electrical and mechanical systems for upgrades, routine maintenance and repairs. For more information, contact Gary New at ext. 1212, garyn@ucar.edu ============================================================================== ============================================================================== NCAR/CGD Systems Work Wed 31 Aug 05 6:00am - 8:00am /home, /var/mail offline ------------------------- /home and /var/mail will be offline for a server reboot to install new firmware. NOTE: Unix servers and desktops _may_ need to be rebooted. We won't know until after the upgrade. Bangkok ------- Bangkok will be offline for quota implementation on bangkok:/home. Calgary ------- . Calgary will be offline for quota implementation on calgary:/home . Calgary will be down to phsycially move it to another rack. Eight nodes will be added during the move for a total of 16. /fs/cgd/* --------- /fs/cgd/home0 and /fs/cgd/data0 will be offline for implementation of quotas. Mineral ------- Mineral will be down for the addition of a new RAID unit (adding 3TB of usable space). Linux /usr/local ---------------- /usr/local on Linux systems will be updated. Updates include: . IDL . New PGI license that _may_ fix the multiple checkout problems. ============================================================================== ============================================================================== NCAR/CGD Systems Work 6:00am - 8:00am Wed 17 Aug 05 As usual, if there are questions or a need to reschedule this work, please contact the CGD/IS staff at systems@cgd.ucar.edu. Please make sure all login sessions and e-mail sessions are closed during this period to avoid file corruption, etc. /project/* ---------- The /project servers will be rebooted to increase the number of of NFS threads. . /project/* will be briefly unavailable. . Some systems _may_ need to be rebooted to correct Stale NFS mounts. We'll check those. . Some desktops systems may need to be rebooted to correct Stale NFS mounts. Users can reboot desktop systems (using the login menu) as necessary. /project/(gsp|tss|cdp|ccr) -------------------------- These project areas will be increased from 200 GB each to 400 GB each. The remaining /project/
partitions (cas, cseg, cas, cms) have already been increased. . No down time is expected during this operation. Mineral ------- . Mineral will be rebooted to clear Stale NFS file handles. Performance Monitoring (SARGE on Cheyenne) ------------------------------------------ Cheyenne will be down for a memory upgrade. E-mail Server Down ------------------ The primary e-mail and IMAP server mailhub.cgd.ucar.edu will be down briefly for migration to a new rack. ============================================================================== ============================================================================== Due to the systems work Wednesday morning: Desktop computers accessing the /project/
filesystems will need to be rebooted Wednesday morning. poorman, salina, and possibly neva, flagstaff, goldhill, and sanitas will be rebooted. -------------------- NCAR/CGD Systems Work Wed 10 Aug 05 07:00am - 08:00am /project/
growth project --------------------------------- AFFECTED SERVICES: /project/cms /project/cseg /project/oce /project/cas The CGD Division /project/
filesystems are in the process of being expanded from 200GB to 400GB. Wednesday morning we will take the original 200GB disks offline, and bring online the new 400GB partitions, for the above listed filesystems. ALL USERS SHOULD CLOSE OUT _ALL_ FILES AND PROGRAMS ACCESSING THESE FILESYSTEMS TO AVOID ANY CORRUPTION, **BEFORE** 7 A.M. WEDNES. MORNING. Please notify the CGD/IS at systems@cgd.ucar.edu if this work needs to be rescheduled. Kathy -- =~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~= Kathleen Fischer (303) 497-1644 NCAR - CGD Information Systems kfischer@ucar.edu =~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~= No eternal reward will forgive us now for wasting the dawn. -Jim Morrison ============================================================================== ============================================================================== NCAR/CGD Systems Work 6:00am - 8:00am Wed 3 Aug 05 /data/ ------------------ The last remnants of the /data file system will be removed from CGD systems. This is the final step in a year long campaign to provide better organization to CGD data and a reasonable expansion path. Any files/directories found at /data/ will ONLY be accessible from the /project/ mount point. This will primarily affect web related access. PLEASE change your scripts, web .htmls, etc. BEFORE 8/3/2005 to avoid any access problems! e.g. You will not be able to cd to /data/paleo. Instead you will need to cd to /project/paleo (just used paleo as an example, you can fill in the directory of your choice) The following are mount points this will affect: (the ones on the left are the ones that are going away) /data/acacia --> /project/acacia /data/OCMIP --> :/project/ocmip /data/paleo --> /project/paleo /data/gsp --> /project/gsp /data/ccsm --> /project/ccsmweb /data/southern --> /project/ccsmweb /data/southern/d1 --> /project/ccsmweb/d1 /data/southern/d2 --> /project/ccsmweb/d2 /data/southern/_aux0_ --> /project/ccsmweb/_aux0_ /data/southern/_aux2_ --> /project/ccsmweb/_aux2_ /data/southern/_aux3_ --> /project/ccsmweb/_aux3_ Bangkok, Calgary ---------------- Bangkok and Calgary will be down for round 2 of the ECC memory upgrade. The replacement memory is in, so this time is should work. Klondike -------- Affected partitions: /fs/cgd/csm /fs/cgd/data0 /fs/cgd/home0 /project/satftp1 /project/satftp2 /project/bgc1 /project/bgc2 Klondike will be down for the removal of un-used hardware (disks) and preparation for expansion of /project/
partitions by 200GB over the next 2 scheduled systems works. All Linux Machines ------------------ All linux machines will have their NFS /etc/auto.* files and /etc/fstab updated to reflect the /data changes. ============================================================================== ============================================================================== Subject: UNPLANNED WORK: /fs/cgd/* - Wed 27 Jul 05 Date: Tue, 26 Jul 2005 14:51:57 -0600 From: CGD/IS Organization: NCAR/CGD To: cgd@cgd.ucar.edu UNPLANNED WORK 7:00 - 8:00am Wed 27 Jul 05 /fs/cgd/* --------- /fs/cgd/home0 registered an error in system log files today at 14:01:00. /fs/cgd/home0 will be taken off line for fault diagnosis and repair (if necessary and possible). If may be necessary to reboot systems after this work to clearn "Stale NFS mount" error messages. At this point, no data has been lost. Daily backups of this file system are maintained. If there are any questions, please send e-mail to the CGD/IS systems staff (systems@cgd.ucar.edu). -- Mark --0- ============================================================================== NCAR/CGD Systems Work Wed 20 Jul 05 6:00am - 8:00am Contact the NCAR/CGD/IS staff at systems@cgd.ucar.edu if there are questions or if this work needs to be rescheduled. Salina ------ . Reboot to add more swap space. Calgary ------- . Calgary will be down to upgrade memory to ECC memory Bangkok ------- . Bangkok will be down to upgrade memory to ECC memory on the master node, and node b003, b002, b001. /data/southern /data/ccsm -------------- . These partitions will be moved to new devices and their final resting place in the /project/ccsmweb hierarchy. NOTE: web links may break due to this move. Please check CCSM related web pages after the systems work. NOTE: Several systems may need to be rebooted after this work. We won't know which ones until after the work is complete. -- Mark --0- ---------------------------------------------------------------------- Mark Moore UCAR/NCAR/CGD mmoore@ucar.edu 1850 Table Mesa Drive (W) 303 497-1338 Boulder, CO 80305 (F) 303 497-1324 ============================================================================== Saturday, July 9, 2005 8 a.m. - 1 p.m. OFFLINE: -------- /home, the H drive on PC's, and email access PLEASE log out BEFORE Saturday at 8 a.m. to avoid any file corruption and/or data loss. The server that houses /home and /var/mail will be upgraded to a more robust/reliable operating system. Please let the systems staff know AS SOON AS POSSIBLE if this interferes with any planned work. Kathy ============================================================================== ============================================================================== NCAR/CGD Systems Work Wed 8 Jun 05 06:00am - 08:00am E-mail Server Upgrade --------------------- AFFECTED SERVICES: All e-mail, including webmail. The CGD Division E-mail server will be upgraded with new hardware, new software (sendmail, IMAP & POP daemons, procmail, spam assassin), and a new operating system (Solaris -> Linux). ALL USERS SHOULD CLOSE OUT _ALL_ E-MAIL AND WEB (ie, Netscape) CLIENTS AT C.O.B. (CLOSE OF BUSINESS) TUE 7 JUN 05 TO AVOID CORRUPTING E-MAIL FOLDERS. The new software versions ares expected to surface existing problems that are being hidden by the current system. This include: . broken procmail filters; . excessively large inboxes; . deleted messages reappearing; . duplicate e-mail messages. FAILURE TO CLOSE ALL E-MAIL AND WEB CLIENTS WILL COMPOUND THESE PROBLEMS. Please notify the CGD/IS at systems@cgd.ucar.edu if this work needs to be rescheduled. -- ============================================================================== ============================================================================== NCAR/CGD Systems Work Wed 25 May 06 6:00am - 8:00am Linux /usr/local Cleanup ------------------------ /usr/local on CGD Linux desktops and servers will have several outdated packages deleted. These include: . intel-mkl-6.0p-11 . intel-icc-8.0.055 . mpich-1.2.5-pgi-hpf-cc-4.1-2 . mpich-1.2.5-pgi-hpf-cc-5.0-2a . pgi-hpf-cc-4.0-2 . pgi-hpf-cc-4.1-2 . pgi-hpf-cc-5.0-2a Linux Systems ------------- CGD Linux systems will be set to not dump core on crash. There is a vulnerability in the core dump procedure. Rather than go through (yet) another kernel upgrade, the crash dumps will be turned off. External Access to WREQ Turned Off ---------------------------------- The only WREQ forum currently available outside of CGD are the CCSM logs. This forum is seldom used, and seems to attrack help desk requests from external users. To avoid confusion, external access will be turned off. Webmail ------- The webmail server will be down while the backend server is upgraded. E-mail ------ E-mail will be down for the installation of a new tape robot. ALL USERS SHOULD EXIT E-MAIL CLINET (i.e., Netscape, Thunderbird, etc) TUESDAY NIGHT TO AVOID CORRUPTING E-MAIL. Contact the CGD/IS staff at systems@cgd.ucar.edu if there are any questions. ============================================================================== Subject: SCD Comp Rm Shutdown Thursday 5/19 - Sunday 5/22 Date: Wed, 18 May 2005 13:35:42 -0600 SCD Computer Room Shutdown Thursday 5/19 - Sunday 5/22 In case you haven't heard.... SCD will be shutting down their computer room this weekend for cooling and electrical system work. Some services will be unavailable all weekend, others will only be down for a few hours on Saturday morning. CGD/IS will not be shutting down during this time. The only effects on CGD systems will occur on Saturday during the network outage (see below). CGD servers are expected to recover gracefully once the network is available. If not, we'll be in Saturday afternoon to fix things. Below is a list of the affected services: Shutdown: Thu. 19 May 2005 at 0000 hrs thru Sun. 22 May 2005 *beginning* at 1730 hrs: - Mass store - All supercomputing systems including bluesky, tempest, pegasus, frost, lightning and bluedawn which require high levels of cooling will be unavailable during the shutdown. - UCAR room reservation system ....and: Shutdown: Sat. 21 May 2005 at 0800 hrs thru Sat. 21 May 2005 by 1200 hrs - Networking (i.e., ANY type of network access, including ALL forms of remote access. This includes access to CGD systems, also.) - UCAS Token authentication (OTP via cryptocard or snkey) - VPN access - UCAR Web services - VPN access - PGI license server (All CGD systems will be affected by this.) - Wireless access - Outbound long distance calling will be unavailable - Voicemail UCAR-wide will be unavailable ....and more: Shutdown: Sat. 21 May 2005 at 0800 hrs thru Sat. 21 May 2005 by 1600 hrs - Email relay system - Maia Mailguard website ============================================================================== NCAR/CGD Systems Work Wed 18 May 05 6:00am - 8:00am New E-mail Server ----------------- A new e-mail server will be installed. No changes to user e-mail clients are required. ALL USERS SHOULD EXIT E-MAIL CLIENTS ON TUESDAY NIGHT PRIOR TO THE INSTALLATION TO AVOID MAILBOX CORRUPTION. /fs Migration SCD -> CGD ------------------------- The /fs/cgd/home0, /fs/cgd/data0, and /fs/cgd/csm data partitions will be migrated from SCD servers to CGD servers. A final data sync will be performed before the cut over. USERS SHOULD NOT ACCESS /fs PARTITIONS DURING THE DATA SYNC TO AVOID FILE CORRUPTION. Project Fileserver Down ----------------------- One of the /project file servers will be down for the installation of a new RAID system. The following data areas will be unavailable during this time: /project/satftp1 /project/satftp2 /project/bgc1 /project/bgc2 Linux /usr/local Upgrade ------------------------ The Linux /usr/local will be upgraded. Changes include: . New PGI license files that will access a new (and hopefully more stable) license server. . New FireFox upgrade with security fixes. Goldhill Down ------------- Goldhill will be down for removal of a disk. As usual, please notify the CGD/IS staff (systems@cgd.ucar.edu) if this will cause any serious impact on planned computing. ============================================================================== ============================================================================== Subject: UNPLANNED WORK: /fs/cgd/* - Wed 27 Jul 05 Date: Tue, 26 Jul 2005 14:51:57 -0600 From: CGD/IS Organization: NCAR/CGD To: cgd@cgd.ucar.edu UNPLANNED WORK 7:00 - 8:00am Wed 27 Jul 05 /fs/cgd/* --------- /fs/cgd/home0 registered an error in system log files today at 14:01:00. /fs/cgd/home0 will be taken off line for fault diagnosis and repair (if necessary and possible). If may be necessary to reboot systems after this work to clearn "Stale NFS mount" error messages. At this point, no data has been lost. Daily backups of this file system are maintained. If there are any questions, please send e-mail to the CGD/IS systems staff (systems@cgd.ucar.edu). -- Mark --0- ============================================================================== NCAR/CGD Systems Work Wed 20 Jul 05 6:00am - 8:00am Contact the NCAR/CGD/IS staff at systems@cgd.ucar.edu if there are questions or if this work needs to be rescheduled. Salina ------ . Reboot to add more swap space. Calgary ------- . Calgary will be down to upgrade memory to ECC memory Bangkok ------- . Bangkok will be down to upgrade memory to ECC memory on the master node, and node b003, b002, b001. /data/southern /data/ccsm -------------- . These partitions will be moved to new devices and their final resting place in the /project/ccsmweb hierarchy. NOTE: web links may break due to this move. Please check CCSM related web pages after the systems work. NOTE: Several systems may need to be rebooted after this work. We won't know which ones until after the work is complete. -- Mark --0- ---------------------------------------------------------------------- Mark Moore UCAR/NCAR/CGD mmoore@ucar.edu 1850 Table Mesa Drive (W) 303 497-1338 Boulder, CO 80305 (F) 303 497-1324 ============================================================================== Saturday, July 9, 2005 8 a.m. - 1 p.m. OFFLINE: -------- /home, the H drive on PC's, and email access PLEASE log out BEFORE Saturday at 8 a.m. to avoid any file corruption and/or data loss. The server that houses /home and /var/mail will be upgraded to a more robust/reliable operating system. Please let the systems staff know AS SOON AS POSSIBLE if this interferes with any planned work. Kathy ============================================================================== ============================================================================== NCAR/CGD Systems Work Wed 8 Jun 05 06:00am - 08:00am E-mail Server Upgrade --------------------- AFFECTED SERVICES: All e-mail, including webmail. The CGD Division E-mail server will be upgraded with new hardware, new software (sendmail, IMAP & POP daemons, procmail, spam assassin), and a new operating system (Solaris -> Linux). ALL USERS SHOULD CLOSE OUT _ALL_ E-MAIL AND WEB (ie, Netscape) CLIENTS AT C.O.B. (CLOSE OF BUSINESS) TUE 7 JUN 05 TO AVOID CORRUPTING E-MAIL FOLDERS. The new software versions ares expected to surface existing problems that are being hidden by the current system. This include: . broken procmail filters; . excessively large inboxes; . deleted messages reappearing; . duplicate e-mail messages. FAILURE TO CLOSE ALL E-MAIL AND WEB CLIENTS WILL COMPOUND THESE PROBLEMS. Please notify the CGD/IS at systems@cgd.ucar.edu if this work needs to be rescheduled. -- ============================================================================== ============================================================================== NCAR/CGD Systems Work Wed 25 May 06 6:00am - 8:00am Linux /usr/local Cleanup ------------------------ /usr/local on CGD Linux desktops and servers will have several outdated packages deleted. These include: . intel-mkl-6.0p-11 . intel-icc-8.0.055 . mpich-1.2.5-pgi-hpf-cc-4.1-2 . mpich-1.2.5-pgi-hpf-cc-5.0-2a . pgi-hpf-cc-4.0-2 . pgi-hpf-cc-4.1-2 . pgi-hpf-cc-5.0-2a Linux Systems ------------- CGD Linux systems will be set to not dump core on crash. There is a vulnerability in the core dump procedure. Rather than go through (yet) another kernel upgrade, the crash dumps will be turned off. External Access to WREQ Turned Off ---------------------------------- The only WREQ forum currently available outside of CGD are the CCSM logs. This forum is seldom used, and seems to attrack help desk requests from external users. To avoid confusion, external access will be turned off. Webmail ------- The webmail server will be down while the backend server is upgraded. E-mail ------ E-mail will be down for the installation of a new tape robot. ALL USERS SHOULD EXIT E-MAIL CLINET (i.e., Netscape, Thunderbird, etc) TUESDAY NIGHT TO AVOID CORRUPTING E-MAIL. Contact the CGD/IS staff at systems@cgd.ucar.edu if there are any questions. ============================================================================== Subject: SCD Comp Rm Shutdown Thursday 5/19 - Sunday 5/22 Date: Wed, 18 May 2005 13:35:42 -0600 SCD Computer Room Shutdown Thursday 5/19 - Sunday 5/22 In case you haven't heard.... SCD will be shutting down their computer room this weekend for cooling and electrical system work. Some services will be unavailable all weekend, others will only be down for a few hours on Saturday morning. CGD/IS will not be shutting down during this time. The only effects on CGD systems will occur on Saturday during the network outage (see below). CGD servers are expected to recover gracefully once the network is available. If not, we'll be in Saturday afternoon to fix things. Below is a list of the affected services: Shutdown: Thu. 19 May 2005 at 0000 hrs thru Sun. 22 May 2005 *beginning* at 1730 hrs: - Mass store - All supercomputing systems including bluesky, tempest, pegasus, frost, lightning and bluedawn which require high levels of cooling will be unavailable during the shutdown. - UCAR room reservation system ....and: Shutdown: Sat. 21 May 2005 at 0800 hrs thru Sat. 21 May 2005 by 1200 hrs - Networking (i.e., ANY type of network access, including ALL forms of remote access. This includes access to CGD systems, also.) - UCAS Token authentication (OTP via cryptocard or snkey) - VPN access - UCAR Web services - VPN access - PGI license server (All CGD systems will be affected by this.) - Wireless access - Outbound long distance calling will be unavailable - Voicemail UCAR-wide will be unavailable ....and more: Shutdown: Sat. 21 May 2005 at 0800 hrs thru Sat. 21 May 2005 by 1600 hrs - Email relay system - Maia Mailguard website ============================================================================== NCAR/CGD Systems Work Wed 18 May 05 6:00am - 8:00am New E-mail Server ----------------- A new e-mail server will be installed. No changes to user e-mail clients are required. ALL USERS SHOULD EXIT E-MAIL CLIENTS ON TUESDAY NIGHT PRIOR TO THE INSTALLATION TO AVOID MAILBOX CORRUPTION. /fs Migration SCD -> CGD ------------------------- The /fs/cgd/home0, /fs/cgd/data0, and /fs/cgd/csm data partitions will be migrated from SCD servers to CGD servers. A final data sync will be performed before the cut over. USERS SHOULD NOT ACCESS /fs PARTITIONS DURING THE DATA SYNC TO AVOID FILE CORRUPTION. Project Fileserver Down ----------------------- One of the /project file servers will be down for the installation of a new RAID system. The following data areas will be unavailable during this time: /project/satftp1 /project/satftp2 /project/bgc1 /project/bgc2 Linux /usr/local Upgrade ------------------------ The Linux /usr/local will be upgraded. Changes include: . New PGI license files that will access a new (and hopefully more stable) license server. . New FireFox upgrade with security fixes. Goldhill Down ------------- Goldhill will be down for removal of a disk. As usual, please notify the CGD/IS staff (systems@cgd.ucar.edu) if this will cause any serious impact on planned computing. ==============================================================================