Saturday, August 26, 2006

windows 2003's RAID5 doesn't support hot-spare --- observed with help from the great VMWare Player

I finally found some scsi virtual disk image to be used for my Windows 2003 server, a virtual machine I set up inside the 'VMWare Player"  in prep for 70-290,  my first MCSE test. In the past weeks, I had a hard time to have more than 2 free (IDE) disks in order to simluate an environment where I can set up RAID5 volumes.

The successful count of the RAID-5 capability inside windows 2003 server:
  • lsiLogic is the controller type supported by the Windows 2003 server. busLogic is not supported w/o external driver.
  • run Device manager to remove the old/broken busLogic controller then re-scan for hardware change
  • run Disk Manager (diskmgmt.msc)
    •  initialize all four brand-new identical 500MB scsi disks
    • convert them all into dynamic disks
    • create a RAID-5 volume using 64MB on each disk: right click on one of them, select 'create a volume' from the pop-up menu, check 'RAID-5', select disks & sizes to create.
  • Nowhere, hot-spare is found in the create-a-volume wizard, or the resultant 3-disk RAID5 volume with one free dynamic disk available in the system
No wonder 'VMWare Player' is voted the best technology of the year by some IT pro magzine. Using 'VMWare Player', I just inserted a few entries to the .vmx configuration file claiming I have a few new disks and a SCSI controller now, on a suspended server. One the server is resumed, the controller & the disks are detected and installed properly. Can you imagine the hassle I would have had in a non-virtualized world. Arghhhhhhhhhhhhhh....
  • buy a busLogic SCSI controller, only to find it doesn't work.
  • waste time searching for proper driver, or return it and get a lsiLogic one to try my luck.
  • buy 4 scsi disk to set up 3-disk RAID5, only to find the 4th disk can't be used as hot-spare anyway
  • shutdown os, power off the box, open the case, screw driver, struggling with free dribe bays & cables, etc.
To have four SCSI disks on a SCSI controller, I just add the following to my blah.vmx:
scsi0.virtualDev = "lsilogic"
scsi0:0.present = "TRUE"
scsi0:0.fileName = "sda.vmdk"
scsi0:0.deviceType = "disk"
scsi0:1.present = "TRUE"
scsi0:1.fileName = "sdb.vmdk"
scsi0:1.deviceType = "disk"
scsi0:2.present = "TRUE"
scsi0:2.fileName = "sdc.vmdk"
scsi0:2.deviceType = "disk"
scsi0:3.present = "TRUE"
scsi0:3.fileName = "sdd.vmdk"
scsi0:3.deviceType = "disk"

VMWare player automatically added a few lines to the blah.vmx file after my lines too:
scsi0:0.redo = ""
scsi0:1.redo = ""
scsi0:2.redo = ""
scsi0:3.redo = ""
scsi0.present = "TRUE"

Thursday, August 24, 2006

sp5 for BES v4.0 :: many more fixes than other service packs

I finally got some time to myself today: my son is asleep when I am
back from work. So, I went ahead to download the release-notes for
service pack 5 for BES v4.0. For work, I administrate a BES 4 server
currently on v4.0.4. To my big surprise, there's tons of fixes in
this service pack, many are for bugs for old sp1/sp2 and many are for
serious BES performance/reliability issues. I don't recall this many
fixes in release notes for other service packs. (To be fair, I started
with 4.0.2) I guess this could be that RIM finally can concentrate on
its product and customers, after settling the patent lawsuits.

on-call duties :: to pay or not to pay

at work, we are talking about setting up a duty pager and a group of
us will join the rotation to be on-call 24x7x365. Some of us don't
think, or think the boss won't like, that the rest of us actually want
to be compensated for the on-call duties. To me, it makes sense to be
compensated for the inconvenience of carrying it, and also be
compensated for the excessive time working off-hour by comp-time or
A strong counter argument is that labor laws categorizes most of IT
professionals as exempt, in that we can't bill for hours outside a 8-5
work day. Many employers forget to see/emphasize/respect the other
side of the token. That is, the exempt status would be void if an
employer ducks our pay if we take a few hours off here and there.
Most don't deduct pay, but they frown upon and ask you to make up the
time. As a result, most of us opt not to take random hours off. Plus,
at smaller places, your chores pile up and your projects get delayed.
In the end, your performance (real or in the eyes of the boss-man) is
hurt, so are the salary and upward movement on the corporate ladder,
in turn.
It gets even murkier to us laypersons if on-call duties is actually
part of your job description. Does that entitle an employer to get you
work whole weekend plus a full week? what do you think?

Sunday, August 20, 2006

fedora core 5 -- better SELinux policy

Tonight I created a virtual machine inside my vmware player, with 256M ram and 10G hard disk, to install Fedora Core 5 from the DVD Andy@redhat handed out at the Intel/Redhat breakfast seminar series. The installation wizard has some good selection menus for SELinux set-up.

This reminds me of what I read in FC5's release notes that the targeted policy in RHEL4 (or CentOS 4 or FC4) is now deprecated. Plus the integration of Zen and JBoss, FC5 now began to look like a larger bite. Ken @ worldspan is certified for RHL 8 and he said he will take RHCE once RHEL 5 is out. wonder if I should pair up with him to get my heart-thorn over with.

Sybase 15.0 certification is ready

I received an email from Sybase mid week. Seeing it says 'free
certification voucher', I forwarded to my former co-worker, who is a
Sybase DBA hunting for a job. I thought it'd be a nice addition to
have a renewed certification on the resume, plus a good distraction
against anxiety while staying home.
After sending the email, I click on 'display image' inside gmail. now
I knew it was not "free' free, but free after taking a few prep
courses offered by Sybase for a few grands. Don't I feel like a

Saturday, August 19, 2006

Redhat/Intel breakfast : Linux para-virtualization on Intel dual core processor

This past Tuesday, I attended the breakfast seminar series of Linux Virtualization and Dual Core processors at Westin near I-285 and peachtree-dunwoody interaction. Once the Westin waitress followed me into the breakfast area and lifted several party tray covers, I instantly understood why the invitation email said 'hot breakfast'. The breakfast consists of hot scrambled eggs, steamed sausages, fried bacon, oven-warm bread, pancakes, and ice-chilled fresh orange juice. "Nice", many said.

Ken from WorldSpan sat next to me. He's a programmer turned systems administrator. He told me that the WorldSpan building at I-75 and I-285 I drive by everyday is the headquarters. Its data center is down at the airport and it used to be part of Delta. They have mostly mainframes with code written in the '60s, a few Solaris servers, and a few Linux servers.

The Intel speaker is OK. A few points I take home are:

* AMD has a very very small footprint in terms of market share. Based on the noise it generates in the media, I thought it is making strides.

* Intel's dual Core processors will come hard and heavy to the market in the coming

* Intel led AMD in 20 of the 22 industry benchmark results.

Andy from Redhat presents the Xen in Fedora Core 5 (and the coming
RHEL 5, beta1 at the end of August). A charming Brit. Xen basically
allows Para-Virtualization, with which the guest OS is aware of the
fact that it is a guest and its kernel is also modified to expedite
its access to raw hardware, esp. CPU/memory. One thing it doesn't do
is the bandwidth allocation for guests.
Two cool features I like about Zen (or enhanced feature available via
RHN, Redhat Network, subscriptions) are:
* snapshot of a guest. such a snapshot can then be used to restore or replicate
* a guest can be migrated, alive, to another physical host, with
minimal downtime appreciable to the end users.

I got a FC5 DVD (good thing I didn't start my download a few days ago)
from a pile Andy put out. Ken and I recalled the fact that we used the
3.5" floppy diskettes to install Debian 0.92 in '96 onto a 486dx.

Friday, August 18, 2006

Farewell group lunch at Pappadeaux

I attended a farewell group lunch today at Pappadeaux . The farewell is to a manager who was replaced. He moved down from MD to GA a year ago, after some sweet&heavy courting.

This is the first time I ever entered a Pappadeaux. It is trendy, expensive, crowded at 12:00pm-2:00pm. Feeling a bit cool, I noticed the sweat on several waiters' foreheads. Entrees started at $15 each. No free bread or butter while waiting. Half of us finished, while the other half waited over 45 minutes to get our plates.

I planned for an hour and ended up spending two hours there for a rather simple lunch. It is not like anybody ordered a full course of meals. Most ordered just a salad, a gumbo bowl, or a sea-food plate. Fortunately my boss sat next to me and we discussed some project priorities. So, I count that as 'working' to cut my lunch time back to an hour, the originally allocated time slot. It is Friday after all, who wants to start his/her weekend one hour late?

For a person who religiously brings brown-bag every day and often frets over a $12 dinner entree at one's favorite authentic ethnic restaurant, I ordered a fried shrimp & scallop lunch plate for $14.99. There's hardly anything you can order to get below that price anyway. I know some at the table earn much less than me and have much more obligations than me. Not sure how they feel about the price tag. Everybody looked and acted happy and cheerful.

Near the end of it, the waiter was about to get checks for everybody. The supervisor of the parting manager suddenly said, "let me pay for all." Joy and shriek across the tables. Many regret that they didn't order a lobster to-go :)

Blog is not part of the Web -- overengineering at Google ?

More often than not, when I search for a non-mainstream topic, I couldn't find it in Web or Group using Google search. However, search by 'blog' may get me a boat-load of results. I wonder why Google doesn't integrate blogs into its web search.

Blogs are mainly addition/appending, just like regular web page. Therefore blogs should be friendly enough to lend themselves to the same storage and indexing algorithms Google uses for web contents. Why bother to separate them into two tab? Even worse, 'blog' is now a few clicks away, instead of just another tab. Is there some over-engineering sprouted at Google ?

Citrix Netscaler :: wildcard not so wild for Content Switch (continued)

I dialed into a conference call with the Citrix lady this afternoon. She told me that answers she got internally are that the RegExp pattern "jack*" wont' work at all for any builds for current Netscaler 6.1 release. What a disappointment! Her inquiry on whether the coming 7.0 release, currently in RC-CR phase (RC stands for release candidate while CR stands for controlled release), will have it was redirected to Citrix developers in India. Answer is pending,
since people are sound asleep at that part of the world right now.

Not sure why this would be a question for developers instead of for some project manager here in the states. You'd think there got to be some form of project plan or use cases or feature lists or fixed bugs for a major release now in RC mode? Should I hold that against Citrix's development or PM process?

Could I just read too much into it, as usual? No. This is by stark contrast with F5's much more mature and systematic approach towards release management.

certification or a M.S./MBA

At work, a biz discussion turned into a chit-chat this morning. A bit of surprise to me, my fellow IS engineer has decided to give up his SCJP4 certification prep and started to prep for GMAT instead. Being a CCNP & MCSE, he came to a painful conclusion that the certification path is too tiresome since they expire (in real term, as well as fading in your mind) before you can use it for work or for job hunting.

Over years of turmoils about to pursue certifications or not, I total agree. Unfortunately, I have a MS in MIS from a prominent school. A few years later, I still don't see a door opened for me to the management track. So, I'd say MS doesn't work, since it labels you as a techie more than anything else. A techie needs something else to package/wrap/color himself or herself in order to impress your future employer that he/she is biz/management material. Recently, a few of my techie friends spent a lot (tuition/time/energy) to get their MBA. That seemed to have paid off. All found positions that a techie would never have dreamed of getting.

ethereal renamed to wireshark -- in June 2006

Yesterday, while reading a RHSA (redhat security advisory, I signed up for notification emails) regarding ethereal, I was surprised to find that ethereal was renamed recently to 'wireshark'. not sure why. since it is such a powerful & popular network troubleshooting tool, I imagine there'd be a lot of people updating their resumes to insert '(wireshark)'.

It was back in June that ethreal was renamed, a tool I use on a daily basis (if not hourly). Am I too aloof to what's going on outside my little "circle of life"? Maybe it is about time for me to get onto some RSS feed or radio feed, so I can automatically pick up such vibes w/o conscious efforts.

citrix netscaler :: wildcard not so wild for Content Switch

With my son going to church group meetings, I had a bit free time to resume playing with the citrix Netscaler EE-9000. I set up a policy using in-line expression to say I want " == jack*", then I created a content switch virtual server to associate the policy with a LB virtual server I created earlier. Simple enough, huh? Not really, all attempts to hit the CS-LB service (firefox, IE, curl) were met with "500 service not available error". Ethereal dumps on the client and tcpdump dumps on the Citrix box couldn't give me anything wrong with communication either.

Combing through the ICG (Installation & Configuration guide), I found such an error usually is due to the fact that an underlying feature is not enabled. I went back to the system node on the configuration GUI. And sure enough, 'content switching' is not checked by default under 'Features'. (while writing this, I am thinking, would it be better to shadow it out if such a feature is not enabled!) Checked & saved. Still got the same error. reboot the box, delete/recreate the policy & cs virtual server, all to no avail.

I cried for help by calling up my friend. He in turn got his lady friend, a Citrix SE, to chat with me. Off the bat, she told me that she's young with the company and may not know all the answers. I said to myself, "oh least she's honest." After a few go-to-meeting sessions and hauling other SEs into it, she finally broke it to me, "no, no wildcard would work for content switching." Great, isn't it?!

So, she and I went down the path to make-do with the limit set of operands on the GUI, trying to come up with a compound regular expression to mimic this behavior. None could really works so far. She kept telling me that newer builds won't help since they only fix bugs and won't change how RegExp works. She also told me that this IS how a PERL Reg object works for Citrix Content Switching. I was smiling along, begging to differ as a certified PERL programmer.

At this point, I am somewhat perplexed since it is hard for me to believe it didn't have true regular expression support. My employer opts not to use sub-domains or URL to divide traffic to load-balanced clusters. Instead, the sales/marketing geniuses want the customer to have any URL they want, only to find us engineers and architects scrambling to direct requests to these URL to a proper LB cluster. Sina and Google are using Netscaler in thousands. Does it mean that they don't need wildcards?

Wednesday, August 16, 2006

Employer questionnaire

While stir-frying long beans for dinner, I received a call from a recruiter named Jack. He told me that an employer wants an interview and wants me to fill out some simple Q/A. I could tell that he was disappointed that I was not as excited as he was. Plainly, I asked him to send it my way via email then back to cooking. Barely in time to save my stir-fry long beans from burning into dark sticks.

Not sure why he just sent an email saying the same thing w/o any interaction from me. Human nature, or occupational hazard? After dinner, I retreated to my study and checked my inbox. It turned out that the Q/A is not for HR, but simple technical questions regarding Linux/Solaris operating systems. I answered them. To be sure, I googled to double check myself, before I shoot it back to the recruiter.

Three thoughts:

  • If I know where/how to find the info and can use the info properly, would it be enough ?
  • I am certified on both Linux and Solaris and have years of professional experience with them. However, I haven't touched either for a year. Who am I to remember three -v is needed to get the most verbose output from openSSH? Is this a sign that I should run away from this employer if this is his/her expectation?
  • If I do a 'man ssh' on my Linux Desktop to get that the exact answer is '-vvv' instead of 'a few -v' I initially wrote down, am I cheating ?

took practice exam for MCSE (70-290)

I ran into a link to today. My son is taking his nap thus I have half an hour to forty-five minutes to kill. So, I think, why not. I took the example exam for 70-290. the quiz session won't start within Firefox, just spinning forever. It took me a while to find IE, since I've deleted the shortcuts on the desktop and taskbar, since I rarely use it if not forced.
The test itself is pretty good, I'd say. Not sure whether they'd be this many ntbackup/vss question though. Questions are pretty long and the 2-minute clock is literally ticking in my face! got to read quick.

scsi disk image by QEmu for VMWare Player for a Windows 2003 guest

I need to have at least three spare disks to check out the RAID5 in Windows 2003 server, a guest OS running inside VMWare player with Windows XP as the host OS. I usually use QEMU's qemu-img command to create IDE disk images. However, VMWare player declined to take IDE2, and declined to allow me to use IDE1:0 (the cd-rom) as a disk device.

So, I thought I'd just use SCSI disks then. Not so fast, they say. QEMU doesn't support SCSI disk image. A Romaian hacker reported an altered qemu-img to create SCSI disk image. Unfortunately, his link on 'hack a day' returned a 404 error. not sure how far I want to go to get more than three spare disks for the RAID5 configuration practice needed for my MCP 70-290.

Ignoring the warning about the disks are IDE disks labeled as SCSI disks, I booted up the server. found-new-hardware wizard started. a driver was found but an 'aged' one, thus Windows 2003 server fails to install the new hardware named 'Buslogic multimaster scsi adapater' or something like that. bummer. I guess I'd just give it up for now, and take their words for it?

Sunday, August 13, 2006

useless junk with important information

Finally got some free time today to clean up the junk pile on my desk. For a few weeks, I could barely find enough space to put my hands on the keyboard to type.

Majority in the pile are actually important mails, such as a few statements from brokerages, bank, premium increase of home insurance, termite warranty payment stub, 2006 county tax notification, and on and on. I never can draw a clear line whether I should thread them right away or keep them for a year just in case. If I keep, then I need to sort them now and remember to come back to sort out them for threading or archiving yet again. what a bore.

All these smart people on earth, why nothing has been invented for this mess of junk?! Or the blame falls right on me. That is, I have too many possessions and decide not to give up an inch or thread, depending what it is. If I were able to follow the spirit of Zen, I'd be able to live with myself with much less from this earthy world, I'd be just as happy through the warranty period of this bio-form of life in this cycle.

Friday, August 11, 2006

recruiter & spam

I chatted with a recruiter yesterday regarding some not-so-interesting positions they need to fill. I went with the flow, thinking hard the top three things I'd like to change at my current job which would makes me so happy that I wouldn't want to leave any more. I took it as assisted mediatation in seek of ultimate happiness.
needless to say, she verified my email address so she can send me something. Later in the afternoon around 4pm, she sent an invitation for some online skill testing.
Around 6pm, I received over 30 spam emails. You know the kind, porn and stuff.  She sent me regular email before, so I'd hold responsible this online skill testing server or its aid who has added my email addresses. I have 1 or two spam show up in INBOX, while the rest was caught by the ISP's spam filters. once a year I may see a surge like this in the INBOX.

Wednesday, August 09, 2006

inadvertent load test against Sybase ASE --- what did a Hobbit find

Last few weeks, I have been busy building up NMS for our site using Hobbit Monitor, an open-source NMS suite mimicking Bigbrother. For the most part, Hobbit Monitor (4.2-RC-20060712 w/o all-in-one patch) works out of the box for non-SELINUX server. I did my SELinux tweaks so it works on CentOS 4 w/o disabling SELinux. My job this week is mainly adding/customizing various health-checks, with some trials & errors here and there.
From time to time, for production database server running Sybase ASE 12.5.x on CentOS 4/i386, a few checks complained no data reported by Hobbit client. After a few days, I discerned the pattern. That is, [ports] and NMS servers's [hobbitd] are always bad when [files] [msgs] [procs] went CLEAR (no data, but considered OK by design) or PURPLE (non-reporting). Hobbitd's history page has 'client data' shows various truncation message. A post to the mailing list got answer from the author Henrik himself. I bumped up the size limit on client data/message from the default 256K to 1024K. Lo-and-behold, there are over 9000 TCP TIME_WAIT connections (~= new db connection) to the database engines.
Since TIME_WAIT is 2 x MSL (Maximum Segment lifetime), it means the application servers made over 9000 connections to the database server within a four-minute sliding window. Talking with our application architect turns fruitful. He said he has been contemplating to use pooled connection for some auxiliary processes. The main web applcation counts for only 20~40 connections (TCP ESTABLISHED) to the database since they are pooled. Recently in our performance lab we are testing how scalable Sybase ASE is for our web application. Now we sorta know. It can handle >= 2125 new database connections per minute, or 3 millon a day :)

Tuesday, August 08, 2006

no T-mobile coverage in my cube, or building?

I received a T-mobile Blackberry 8700g demo unit. turned it on in my cube. Surprise surprise, no coverage. walk out still no coverage. wandered around got no coverage. until a Blackberry scout told me to lean the unit against the north glass wall. That's pretty pathetic, consider that 1) the company is a T-mobile partner and this unit is one of the many demo units from T-mobile 2) the building is located less than 5 mile north of I-285 and I-75 intersection.

Monday, August 07, 2006

Thunderstorm and desktop UPS and RMA policy [revised]

Saturday night we had decent thunderstorm in the area. It killed the battery-protected section on a Energizer desktop UPS. In the past two years, thunderstorm always killed TrippLite UPSes in the house. To its credit, TrippLite is quick to ship out RMA unit and doesn't care about return the broken one once confirmed it is bad.

One Belkin 350VA was killed too last December, along with a 900MHZ cordless phone connected to it. Belkin did offer me RMA only for the unit. Their rep insisted to have the bad unit returned first and warned me that the replacement unit would have to be a Belkin 500VA with 6 sockets. The dead Belkin 350VA had 8 sockets. I can barely make do with 8 scokets, so I argued for a replacement unit with at least 8 sockets. The anwer is a resounding no. Plus, It may cost me over $10 to ship the heavy sucker back. So, it sits in my fireplace along with TrippLittes.

I couldn't help but wonder whether this is a conspiracy to deter RMA ?! To me, the manufacturer should bear the cost of two-way shipping for RMA. As their paying customers, we are gracious enough not to charge them for the hassle and loss due to ineffectiveness (or defect) in their product, or handling fee!

I guess I'd need to call Energizer to see what's their RMA policy is. It definitely sucks to have the battery section of the sockets stop working. If the thunderstorm can kill UPS itself, I can't imagine what would have happened if the computers/cable modem/VoIP adapter or HDTV connected directly to the wall outlets. The house was built in mid '01, so I don't think wire age comes to play here.

Sunday, August 06, 2006

software RAID capability is Nice in Windows 2003

I have years of hands-on experience with Soltice Disk Suite under Solaris and LVM or alike under Linux. Trying hard to stay open-minded with ears up over the years, I don't think I heard a beep about the nice software RAID capability inside Windows 2003. Don't have figures for reliability/performance and such in the self-training kit book for 70-290. That aside, I was pleasantly surprised and truly impressed by the wizard driven RAID capability. Using Disk Management ( diskmgmt.msc), one can convert a basic disk to a dynamic disk, then to create a new volume by extending (spanning), stripping [RAID 0], mirroring [RAID 1], stripping with parity [RAID 5].

how much should I pay for a RJ-11 duplexor?

While at the brand-new Super Wal-Mart, I bought a phone jack duplexor for $4.24 plus tax. A Belkin one is listed at $0.88 at HP's website with $5 shipping, while ~$6 at RadioShack's website. I need it to have my desk phone work again whose jack has been recently taken by my new VoIP adapter. wonder why my VoIP vendor doesn't supply a duplexor. Liability concern ? Techies I know all plug their VoIP adapter's FXO port to a regular phone jack in the house thus all the jacks in the house are hot again, after disconnecting the regular phone line off the hook at the curb.

super wal-mart moved in, when'd Fry's ?

a super wal-mart opened this week three miles from our house. It replaced a regular Wal-Mart some snobbish neighbors had complained about.

We toured it today. Walking the long aisles packed with snacks and such, I realized a trend. That is, Wal-Mart has determined to move closer to the city, to the rich neighborhoods which Wal-Mart used to shun away from due to consideration of target clienette, as well as real estate price. Does this mean the biz at far-out skirts are saturated, and/or the real estate price is comparatively reasonable enough near the city now. Or, Wal-Mart just wants a piece of the juicy/fat pie that Costo is having esp. during recent econoic downturn?

No matter the answer, the competition is definitely ON, for Costco, BJ, Target/Kroger/Publix and other mom-pop shops. As a techie, I can't help but wonder when Fry's would follow the lead ? Right now, the closest Fry's is 20 miles away.

Saturday, August 05, 2006

Hobbit monitor :: a nice BigBrother replacement

it took me a bit to hunt down the good community of Hobbits, and it is well worthy. Hobbit Monitor's site is not kept up-2-date. Wonder whether the wrong impression people get off the SF site misled people to believe it is not maintained up-to-date. The Hobbit Monitor seems to be nice & functional Big Brother replacement. The active development and the author responds to mailing-list question and code/attach patch while discussing a bug with an end-user, is reminiscent of the early open-source days or early Internet days. I definitely like the ACK feature and the zoom-in on history and trend graph (guess it'd be part of the rrdtool function as well).

Friday, August 04, 2006

Bottom of your career: replacement is being introduced to co-worker across the hallway

Today at work. Sitting in my cube & minding my own biz, I was interrupted by a director for another group. He introduced a new hire to me. The new gal's title is blah1 team lead. I was just about to say, innocently, 'wasn't this the same title of John Unlucky' across the hallway, when the "friendly" director said, "John is leaving."

Hmmm... What do you learn from this? What should I learn? So true it is a two-way street in the corporate America. You can fire the boss at will, so can he/she. If it is so true, why am I somewhat upset by it? Did I expect a month notice before firing, or some second chances? Am I upset because I suspect that I may appear to be a dispensable as well. Or, am I upset only because it is not a good business practice nor humane practice? The first let-go guy is still in the market, three months later, trying hard to explain away why he was let go to employers impressed by the depth of his expertise on the subject matter.

Should Mr. Unlucky celebrate his release from this firm, or cry over his let-go? Given that this replacement-shows-up-first thingy had happened once already, only three months ago, plus the fact that this firm has less than forty staffers, including the CEO, the CFO, and the janitor.

Or, should we just move on and forget about it, as one of my favorite character said once, "who cares? An emperor or a beggar, we are all here to make a living. All is cool as long as we play the part in the show so to make the show real, yet remind yourself it is merely a show."

Citrix Netscaler NS9000 :: DSR mode server selection based on first packet

it is somewhat crazy for a content switch or application switch or any modern web traffic director to make decision based on first request packet. A content switching load balancer definitely needs to get at least all the headers for a HTTP or SMTP request.

It is my experience that there are quite some proxy servers out there who inserted all these non-sense headers, which push the required and essential HOST header for a HTTP/1.1 request to the very bottom of the header section. For one thing, http request from a Blackberry with BIS or BES account will have a huge Accept-type HTTP header, which pushed the Host: header to the very bottom and outside of first packet for a simple HTTP request. As a result, the Host header shows up in second packet or later. In turn, the server selection based on first packet treats this as no an empty Host header and fails, with the client traffic wrongly directed.

It is problematic with anything other than waiting for all headers or at least till the HOST header shows up or till the all HEADERS you have defined to switch directions against. The latter is not fool-proof, since HTTP doesn't prevent you to have the same header over and over again.

Of course, it'd be even crazier if Citrix Netscaler under non-DSR mode does the same thing. I'll report back.

windows 2003 server std (sp1) : disk manager couldn't count ?!

Studying storage chapter for 70-290 last night, I created four 100M virtual disks under VMware Player. IDE2:0 didn't show after reboot, so I guess the free vmware player supports only 4 IDE devices after all.
Using disk manager (diskmgmt.exe), I initialized the three new disks as 'basic disk'. I then tried to create three primary partitions on disk1 in the size of 8M, 16M, 24M and in that order. The first 8M created properly. The 16M one showed up as a 8M, even if the partition-creation wizard's confirmation summary clearly showed I asked for 16M. The 24M one I intended to create was created at11M ?! I wonder if this has anything to do with the fact the minimum size is 8M. For all three, I answered it not to assign drive letter nor to format. Not sure what's wrong with this picture, except maybe, maybe vmware player.

sybase ase hourly transaction log dump keep growing in size (continued)

To me, the replication transactions should not be considered part of normal db transactions, in the sense that they don't alter the data state of the database, but that of the replicated twins. At least, no reason for 'truncation point' for the real db to get held hostage if a gone-wild REP thread declined/forgot/neglected to commit its replication transaction.

Thursday, August 03, 2006

sybase ase hourly transaction log dump keep growing in size

our dba asked me for more space on database servers since it ran out. I opened my trending graph by MRTG. sure enough, the file system holding the backups has creeping up for three weeks now. however, it grows too fast, something like 80G a day instead of 8G a day I'd expect. with the help of 'du' and 'sort', I found the hourly transaction dump is the culprit. It has grown from 10M each to 1G each in the course of three weeks. reported this back to the sybase DBA. after some digging, he said it is a gone-wild replication thread holding a transaction open, somehow. even nightly full dump failed to mark the db such that hourly dump would cover diff instead of accumulative...

Wednesday, August 02, 2006

what a shame: ASR stopped to work (server 2003 SP1 bug & fix)

It is disheartening to have ASR failure after SP1 applied to a DC based on Windows 2003 server. More so is the fix per MS technet article. One option is to relocate the AD/DS log to C:\ where the AD/DS database is. This option requires you to reboot the domain controller server to 'AD recovery mode' then run a dsutils.exe to relocate the logs. This dsutils.exe is not even installed by default! Moreover, having DS log on a different volume is recommended by MS for best-practice. Talking about testing against your own use cases!!!

As I know better with extensive UNIX/Linux experience, such a log relocation would be and should be simply a 'kill -HUP pidOfProcess' after changing some configuration file for this process, no matter how important it is, to point to the new location of the log files. Users should not even know a thing about it.

is this just blastings from a unix guy, or windows guys kinda agree or wished for as well?

deprecated SUS in MCP 70-290 ?

I am prep for my first Microsft certification exam, 70-290. I need to learn this stuff to feel comfortable dealing with the MS 2003 Server which BES 4 is running on. For the past few months, I have been studing with Microsoft's own self-training kit for MCSE 2003. When I googled for 'SUS' to download, a MS link says SUS is deprecated late December. My question is, if I am going to take the test late December, should I study for the new Wuss instead of the SUS ?

RIM released SP5 for BES v4.0

I received an email yesterday from RIM saying SP5 is now available (at least at the partnerzone) for BES (Blackberry Enterprise Solution) v4.0 for Exchange. From my limited 1y+ experience, I have a inkling that their release process is not as stringent as I'd like, since many times I am still surprised to find a disappearing bug or a new bug or two, even after examining their release notes. Regardless, got to go with the flow and read sp5's release notes, since no better alternative is available.

netscaler EE9000 based on Freebsd 4.9

courtesy of a friend, I borrowed a Netscaler (now Citrix) EE9000 load balancer. An 80G SATA drive at the rear end of the pizza-box and a 256M CompactFlash card too. pretty amusing the 'quick-start guide' doesn't say these network configuration is targeted at which network socket. got to admit, it was quick glance.

connected the enclosed null modem serial cable (DB-9 to DB-9) to a nearby Linux boxen. power it up, logon, and saw 'uname -a' as below:
FreeBSD ns 4.9-NETSCALER-6.1 FreeBSD 4.9-NETSCALER-6.1 .

This motivates me to search for pre-made image for VMware player. unfortunately, most sites only have FreeBSD 6.0 or 6.1. wonder if the difference worthy to do the download/install myself. I mainly want to have FreeBSD installation to test disaster recovery or read some man pages, and some 'dangerous' tricks a good system engineer won't attempt on a production server.

hello, Blogger

Doing many different things at work. Doing many different things at home. Doing many different things in between. A highly educated man I am, I want to be reasonably good at everything I do and have made decent efforts to be so. Years passed. In the rear mirror, found myself a technological generalist, a modern-day "jack-of-all-trades".