Differences

This shows you the differences between two versions of the page.

--- haas:status [2015/05/03 10:23] – [new router] wedge
+++ haas:status [2022/08/28 16:15] (current) – [August 28th, 2022] wedge
@@ Line 5: / Line 5: @@
   * update grade not-z scripts to handle/be aware of winter terms
   * update system page for (new)www
-    * include information on db server
-  * upgrade mail VM to jessie/opensmtpd
-  * upgrade LDAP VMs to jessie
-  * create new router, using OpenBSD 5.7
   * scripts to write:
     * user homedir backup pruning script
@@ Line 82: / Line 78: @@
     * https://github.com/zeoeng/zeo-android-api
     * http://eric-blue.com/2013/06/09/life-beyond-zeo/
+  * rxvt / .Xresources
+    * http://www.askapache.com/linux/rxvt-xresources.html
 ======Other Days======
-======May 2nd, 2015======
+======August 28th, 2022======
-=====OpenBSD 5.7 bootable USB creation=====
+=====updated dokuwiki=====
-I finally got around to setting up a bootable USB thumb drive with the OpenBSD 5.7 installer. Quite useful instructions were found at:
+After enduring constant nags of available updates that I could never resolve, and despite an active semester underway, I committed to updating to the recent major release of dokuwiki (igor), and after wrestling with various file permission issues, was seemingly successful in performing the upgrade AND eliminating the annoying nag messages.
+=====dokuwiki included pages no edit button (fixed)=====
+I wanted a read-only base project page, with notes namespace-included content that users could edit.
-  * http://undeadly.org/cgi?action=article&sid=20140225072408
+Unfortunately, that particular combination did not work, at least, not until I did some searching and found: https://github.com/dokufreaks/plugin-include/issues/132
-To summarize, first we "local loopback" attach and mount the 5.7 downloaded ISO:
+<code php>
+function html_secedit($text,$show=true){
+    global $INFO;
-<cli>
+    if((isset($INFO) && !$INFO['writable']) || !$show || (isset($INFO) && $INFO['rev'])){
-# vnconfig vnd0 install57.iso
+        $include_SEC_EDIT_PATTERN = '#<!-- EDIT{(?!.*PLUGIN_INCLUDE).*?} -->#';
-# mount /dev/vnd0a /mnt
+        return preg_replace($include_SEC_EDIT_PATTERN,'',$text);
-</cli>
+    }
-Next, we created the destination image (will will be written to the flash drive). The ISO was 219MB, I made the image 256MB, although we could easily have cut it far closer to 220MB):
+    return preg_replace_callback(SEC_EDIT_PATTERN,
+                'html_secedit_button', $text);
+}
+</code>
+======March 12th, 2020======
-<cli>
+=====borked journal, maybe fix=====
-# dd if=/dev/zero of=disk.img bs=1m count=256
-</cli>
-Local loopback attach our newly created image file:
+<code>
+tune2sf -O ^has_journal /dev/<dev>
+e2fsck -f /dev/<dev>
+tune2sf -j /dev/<dev>
+</code>
-<cli>
+Obtained from:
-# vnconfig vnd1 disk.img
-</cli>
-Partition, slice, and format the disk image in accordance with OpenBSD preferences (I rather enjoyed the echo line automating the disklabel session... another reason why I love that "everything is a file"):
+  * https://www.linuxquestions.org/questions/linux-newbie-8/aborted-journal-and-volume-remounted-read-only-812216/
+======November 21st, 2019======
+=====pi4b dual head=====
+Apparently I was missing something when building up from the lite version of raspbian. For when I popped in a new raspbian (the full desktop version) everything came up just fine.
-<cli>
+Oh well. At least things are now working as they should.
-# fdisk -i vnd1
-# echo "a\n\n\n\n\nw\nq\n" | disklabel -E vnd1
-# newfs /dev/rvnd1a
-</cli>
-And mount it:
+======November 19th, 2019======
+=====pi4b dual head issues=====
+Attempting to set up a fresh raspbian install on the pi4b. Strangely, getting issues with dual head: it is only coming up in mirrored mode (or when I play with it, not at all).
-<cli>
+The only major variables that have changed:
-# mkdir /mnt2
-# mount /dev/vnd1a /mnt2
-</cli>
-Copy the installer:
+  * new raspbian release (I could test my old setup.. that is easy enough to verify)
+  * no longer using HDMI monitors, but the DVI connections on the LAIR monitors.
+    * further investigation suggests the pi may be seeing identical EDIDs and other monitor ID things, so it is defaulting to a mirroring mode.
+    * **tvservice -l** consistently identifies two displays, so the hardware sees it, the software is not.
-<cli>
+Interestingly: my old raspbian setup IS working properly. So what is different between the two?
-# cp -r /mnt/* /mnt2/
+======March 15th, 2018======
-</cli>
+Setting up the raspberry pi 3 for use in the comporg class. Named it **pi3b**, and will get the IP of 10.80.2.48; running the latest stable release of rasbian.
-Make it bootable (this method only applicable for OpenBSD 5.5 and newer, previous versions had at least 2 steps):
+Looking to get LDAP and autofs going, then installing packages to make it minimally functional for our purposes.
-<cli>
+======March 6th, 2018======
-installboot -r /mnt2 vnd1 /usr/mdec/biosboot /usr/mdec/boot
+Aside from an attempt at restarting my status-reporting habits in January of 2017, I totally neglected it for the remainder of the year.
-</cli>
+So here we go, attempting another go. Will we be successful?
-Penultimately, wrap everything up (unmount and detach):
+=====pods configured for better SSD operation=====
+In other recent projects, I encountered some useful configuration options to improve performance, and avoid some anti-longevity side-effects of default settings.
+I ended up taking a Debian USB installer, booting each pod in rescue mode, getting to a prompt, and running the following on EACH:
 <cli>
-# umount /mnt
+# umount /
-# umount /mnt2
+# tune2fs -O ^has_journal /dev/sda1
-# vnconfig -u vnd1
+# tune2fs -o discard /dev/sda1
-# vnconfig -u vnd0
 </cli>
-Finally, actually write it (I did this on a Linux box, as my OpenBSD box is on a rather raw VM):
+Then on the master nodes, I enhanced both **/etc/fstab** and **/etc/rc.local** as follows:
-<cli>
+<code>
-# dd if=disk.img of=/dev/sdb bs=1M
+# /etc/fstab: static file system information.
-</cli>
+#
+# <file system> <mount point>   <type>  <options>         <dump>  <pass>
+/dev/sda1       /               ext4    noatime,nodiratime,discard,commit=600,errors=remount-ro 0       1
+tmpfs           /tmp            tmpfs   defaults,noatime,mode=1777 0 0
+tmpfs           /var/log        tmpfs   defaults,noatime,mode=1777 0 0
+tmpfs           /var/spool      tmpfs   defaults,noatime,mode=1777 0 0
+tmpfs           /var/tmp        tmpfs   defaults,noatime,mode=1777 0 0
+</code>
+Note how I put the log files in RAM too (on the pods)... this should greatly reduce extraneous writes.
+Then at the bottom of **/etc/rc.local**, I added:
+<code>
+#!/bin/sh -e
+#
+# rc.local
+#
+IPADDR=
+while [ -z "${IPADDR}" ]; do
+        IPADDR=`/usr/local/bin/iface | /bin/grep -v 'lo' | /usr/bin/head | /usr/bin/cut -d' ' -f2`
+        sleep 1
+done
+/bin/hostname `/usr/bin/host ${IPADDR} | /usr/bin/cut -d' ' -f5 | /bin/sed s/\.$//`
+/usr/bin/host ${IPADDR} | /usr/bin/cut -d' ' -f5 | /bin/sed s/\.$//  > /etc/hostname
+echo "${IPADDR}    `cat /etc/hostname`"                              > /etc/hosts
+echo "127.0.0.1    localhost"                                       >> /etc/hosts
+echo "10.80.2.46   lab46.g7n.org lab46.corning-cc.edu"              >> /etc/hosts
+echo deadline > /sys/block/sda/queue/scheduler
+echo 1 > /sys/block/sda/queue/iosched/fifo_batch
+exit 0
+</code>
+As the **deadline** scheduler has frequent mention of usage with SSDs.
+Useful URLs:
-I recalled some vague and increasingly distant thoughts over how difficult this once seemed... raw write an image to a device? That needed special tools... nope, just do a low-level binary write... everything is a file.
+  * [[https://github.com/alghanmi/ubuntu-desktop_setup/wiki/SSD-Configuration-on-Linux]]
+  * [[https://www.howtogeek.com/62761/how-to-tweak-your-ssd-in-ubuntu-for-better-performance/]]
+  * [[https://wiki.debian.org/SSDOptimization]]
-Should be good to go.
-======May 1st, 2015======
-=====auth2/auth3 ldap package hold=====
-To avoid manually saying 'n' everyday, I have placed the upgraded ldap packages on hold on those two machines (I suspect auth1 is similar).
-=====grub updates on file servers=====
-I noticed there was grub-level config support for specifying sysvinit instead of the default (systemd), so I specified sysvinit. Rebooted data2 a couple times to make sure everything was working as it should.
-=====pruned backups=====
+=====lab46too SSD updates=====
-I cleared out upwards to 100GB of no-longer-needed user homedir backups. I really should deploy an auto-pruning script once again.
+With the pods having seen their SSD updates, I am looking to do the same for lab46too, testing to make sure the similar changes can be applied to that machine before rolling them over to lab46 proper.
-=====new router=====
+I've updated the /etc/fstab and /etc/rc.local files accordingly on lab46, so those changes have been synced over to lab46too on a recent rsync run.
-One of my summer projects, especially with the anticipated LAIR move, is to update the routers... possibly into just one box, on one of the i5 boxes donated recently. Using OpenBSD 5.7, and running the system out of RAM, much in the same way data1/data2 operate (might as well, I've got enough RAM in it).
+I had to remove the commit=600 from the /home partition. For some reason it didn't want to mount with that.
 ======Month Status Page Navigation======
-^  [[haas/status/status_201504|Previous Month]]  ^  [[haas/status|Current Month]]  ^  [[haas/status/status_201506|Next Month]]  |
+^  [[haas/status/status_201701|Previous Month]]  ^  [[haas/status|Current Month]]  ^  [[haas/status/status_201804|Next Month]]  |