Recently discovered, that it is possible to monitor "Smart Array" controllers under solaris 10 OS. You need to install HPQacucli solaris package, which can be downloaded from hp.com site, read carefully "RELEASE NOTES" about support of your server hardware and controllers, also make sure that you have a recent version of CPQary3 driver installed (2.1.0 or later). Here's an example output:
Going to describe this entry a bit later, but in short "ubuntu server edition can't be plain installed on a server via serial console connection, but with custom made iso/cdrom, which is not always an option".
don't discuss this entry here, better go to bug-report URL and put your thoughts there.
Recently installed a new backup server with Bacula and connected to our SAN based on Brocade switches and HDS (DF700XS unit type, microprogram rev. 0781/A-X). We've never tested this HDS unit, but I was interested, so pasting results here.
I've learnt that solaris 10 x86 doesn't have HME driver. This driver is used for example for older SUN made ethernet cards, and absence of driver for x86 makes them sparc-only hardware (on solaris 10 OS). Linux have driver for HME cards for ages (./linux-sources/drivers/net/sunhme.c comments dated back to 2000 year), both for x86 and sparc.
vzctl is fauting to create container with the following error
Creating container private area (....)
tar: ./boot: Cannot mkdir: Disk quota exceeded
vps-create ERROR: Error in tar -z -xf /vz/template/cache/....tar.gz
Creation of container private area failed
Brought Belkin Wireless G USB Adapter this week. My home ubuntu 8.10 recognised it and networked without any additional settings besides wireless security. Just to let you know.
My new work is to manage solaris 10 on sparc machines, and i'm newbie in solaris.
here's some hints on packages, how to compile something if you installed Solaris in minimal CORE package configuration:
SUNWbtool for ar , SUNWsprot for make , add /usr/ccs/bin to PATH
SUNWhea for std headers (stdio.h, etc...)
last time, our developers have given me following error, while trying to use mysql database connection from php :
File 'NONEXISTENT/charsets/?.conf' not found (Errcode: 2)
Character set '#51' is not a compiled character set and is not specified in the 'NONEXISTENT/charsets/Index' file
i spend 3 working hours to detect what was the reason of this, and the problem was inside php which was compiled with it's internal mysql support. once i recompiled php (--with-mysql=/path/to/mysql) with system installed mysql version, the problem is gone!
we still using old apache version, 1.3.37 to be exact. And today i discovered in its log file:
[Wed Dec 27 18:14:42 2006] [notice] child pid 19870 exit signal Segmentation fault (11)
but i don't found core file to be created. So, i've added CoreDumpDirectory into httpd.conf, but to my surprise, this doesn't help also. googlin a bit, i've found mod_prctl.c which finally helped me t
it is not possible to install oracle on rhel4 server using fedora5 xorg server over ssh forwarded X11 connection, since they are different versions. really, i'm getting blank java window of oracle installer on my fedora5. and if i run installer locally on rhel4 or try old non-secure X11 network connection (export DISPLAY=fc5:0.0) or using port forward over ssh (not x11 forwarding) - it goes just fine. i wander what changed so much in xorg (rhel4 using xorg-x11-libs-6.8.2-1.EL and fedora5 X Window System Version 7.0.0 [xorg-x11-server-Xorg-1.0.1-9.fc5]). and yes, xclock runs well.
Using firefox browser which comes with my Fedora 5 installation, after some time, lets see how much it can "eat" your computer memory: it have 8 opened tabs, from which only one is "complex" (gmail.com)
$ ps axu | grep firefox mator 2823 1.7 37.3 796336 383508 ? Sl Aug18 102:07 /usr/lib/firefox-188.8.131.52/firefox-bin
so, after 4 days of running it taken 800Mb !!! A good way to make firefox to eat all memory is to leave it in memory for 2-3 days with some pages which auto-reload, like gmail, nagios service tracking or some news site.
... jumping through my spam folder today morning, found at least 3 messages marked with SA (spamassassin) :
From: email@example.com To: firstname.lastname@example.org Subject: [SpamCop (184.108.40.206) id:1883243462]Новое сообщение в форуме SEMICVET.RU [!! SPAM] X-Spam-Status: Yes, score=7.2 required=5.0 tests=DATE_IN_PAST_06_12, FORGED_MUA_MOZILLA,FORGED_RCVD_HELO,FROM_STARTS_WITH_NUMS,KAS_SPAM,
using an old celeron 466MHz machine with 2 new Maxtor 6Y080L0 disks in raid1 software array. DMA and 32bit-access is enabled (hdparm) for both disks.
# uptime 17:43:19 up 2:09, 3 users, load average: 1.29, 1.06, 1.06 # mdadm -D /dev/md0 /dev/md0: Version : 00.90.01 Creation Time : Mon Jul 24 12:35:51 2006 Raid Level : raid1 Array Size : 72685952 (69.32 GiB 74.43 GB)
Today was expecting mysql server with average 50 queries per second. Version of mysql server installed is 4.0.24. Found that query_cache_size is default to 0 so it's switched off. Changed /etc/my.cnf for a new size and noticed good cache hit ratio. Too bad this server is only 1Gb of ram, so i can't increase read_buffer_size/read_rnd_buffer_size to bigger values. Also seen too much of 'Opened tables', the value was 10222, so increased 'table_cache' from defaults.