Budak kecik main ayaq acaq,
Perangai buruk mulut loqlaq,
Kalau dah raya memanglah hingaq,
Bagi duit raya tangan dok ketaq.
Lampu raya mengerlip lip lap,
Jamu kueh raya dengan 7-up,
Kalau ada salah dan silap,
Aku minta ampun minta maap.
Selamat Hari Raya Aidil Fitri, Maaf Zahir dan Batin.
* Winthrop Hall's image cilok dari Hamdan UWA. TQ Bro.
Banyak hari dah tak update, so here goes.
Since Monday lepas, aku disewakan contracted to one of the commercial bank kat sini to do some work. Initial timeline katanya 4 weeks. Tapi mungkin lebih or mungkin kurang. Tengok progress aku camna...lembap ker tak. :)
First day aku kat bank tuh, memang sangap. Sebab serba serbi tak leh buat. Satu system pun aku takleh nak access lagi. Kena baca procedures lah, standards lah, etc. Duduk pula dalam cubicles. YM tak boleh, yahoo mail tak boleh, gmail tak boleh, firefox tak boleh. Menguap saja dari pagi sampai petang.
Tapi on Tuesday, dah start kerja keras. Ada 3 projects yang aku kena siapkan. Aku rasa macam tak sempat 3 projects in 4 weeks. First project, kena deploy system management tools nak manage several Linux servers kat sini. Manage patching dan bagai. Nampak macam senang tapi memang gila depa ni. Production servers 5-6 biji tapi tak pernah patch sejak berkurun, lepas tuh nak suruh upgrade OS and patch to latest level. Upgrade OS and patch takdak masalah, tapi abis segala applications yang ntah apa2 dok run kat situ camna nak buat? Dah la tu, servers semua tak leh access internet...local LAN saja. Waduhhh!
Aku dok pikiaq. Selama ni takdak system admin kaaa yang dok maintain servers ni semua? Rupanya dulu memang ada Linux admin tapi dah berambos, yang tinggal cuma AIX admins 3 orang. Dan depa 3 orang ni tak ambik port langsung pasal Linux servers ni.
Antara solutions yang aku jumpa & tanya orang (saja aku letak sini, lain kali boleh buat reference)...
* Red Hat Network satellite server
* Red Hat Network proxy server
* cfengine
* bcfg2
* automateit
* puppet
* func
* Ubuntu Landscape
* Novell ZENworks Linux Management
Setakat ni, potential candidates for solution guna RHN proxy server ataupun puppet. Tapi lom decide lagi.
Second project, deploy openLDAP on several AIX LPARs, dalam environment campur sekali dengan Linux, Solaris and Window clients. Ni aku tak tau camna nak buat. Tapi rasanya boleh buat. Nanti pikiaq camna.
Third and last project pula, clone atau save redundant VIOS configuration files for failover. Yang ni pun aku tak tau camna nak buat. Setahu aku benda ni memang takdak available tools, kena start from scrath. Macam nak kena scripting saja ni. Kalau scripting, memang aku fail. Harap-haraplah tak kena buat project ni. Huhuhu.
Aku rasa bank ni mesti kena pau kaw-kaw dari boss aku untuk "Professional Service" for 4 weeks ni. Angin-angin sepoi yang aku dengar, per day rate is AUD$1000. Nasiblah, sapa suruh jadi bank..kenalah pau sebab duit banyak. Kalau kedai beskal kaa, kedai jual murukku kaa sure tak kena pau melampau. Hehehe.
These are a few sccli commands that I usually use to check on the 3510 disk array,
show media-check
- if any are running, stop them with "abort media-check"show fru
- check for any failed FRUshow disks
- check for any problems with logical drivesshow events
- display eventsshow cache
- confirm write cache is enabledshow ip-address
- show IP addressshow configuration
- show all of the configuration
We can also telnet directly to the 3510 and use a menu-driven command to do stuff on it. Usually once you first connect through telnet, you need to refresh it with Control+L and choose vt100 as terminal type.
To replace a failed disk in the array, first check which disk is failing with sccli :show disks
Or using the 3510 menu :
Main Menu > view and edit Drives
The failed disk most likely will have the status as BAD or FAILED. Take note of the Channel number, target ID column and LG_DRV column. If you already know how the disks are physically placed, you can now replace the failed disk basing on the Channel number and target ID details.
But if you are uncertain the physical location of the failed disk (trust me it's not fun at all when your data got corrupted or lost just because the system engineer pull out the wrong disk from the array!), do from the menu :
Main Menu > view and edit Drives > Identifying SCSI drive > Flash all but selected drive
At this moment, all the LED disks in the array should be flashing except the one that you have chosen (failed disk). Take note of the location and simply proceed to replace the disk. It's how swappable so not need downtime whatsoever.
Once you have replaced tha faulty disk, simply rebuild the RAID from sccli,rebuild idX
Or from menu,
Main Menu > view and edit Drives > rebuild
Simplest one :# /opt/ignite/bin/make_tape_recovery -Av -a /dev/rmt/0mn
Best one :# /opt/ignite/bin/make_tape_recovery -I -v -x inc_entire=vg00 -a /dev/rmt/0mn
1) Getting the default IP address for HMC from the FSP (Flexible Service Processor).
a) At the front LED panel, press the enter button (usually the blue button in the middle) till you get LED code 02.
b) Then change from Normal mode to Manual mode :A < N to A < M
b) Once in Manual mode, go to LED code 30. The HMC address will be at :code 3000 - HMC port 1 - 192.168.255.254
code 3001 - HMC port 2 - 192.168.3.147
You can't access to the LED code 30 if you are in Normal mode that's why you have to change it to Manual mode.
2) Checking if HMC is connected to the server or not.
From FSP LED code 01,HMC = 1, there is HMC connected to the server
HMC = 0, no HMC detected.
3) If HMC is connected, you can't use the serial cable to access the server. In order to use the serial cable, you have to either reset it through ASMI
-> System Service Aids -> Reset Service Processor
or you can remove the FSP card and the battery from the system board to clear the data. Be aware that by resetting the FSP, your LPAR configurations may be lost or corrupted.
4) To connect to the server using ASMI, get a crossover cable and connect it to the HMC port. Set the IP range for your laptop the same as the default IP of the HMC port, depending on which port you use. Open up the web browser and browse to https://192.168.255.254 or https://192.168.3.147
You can also access the ASMI from the HMC, click on the server name then choose Operations > Advanced System Management. The default username and password is admin/admin.
Tunggu taxi dari pukul 5:40 ptg sampai 7:30 ptg, sambil berdiri kesejukan around 11 degree celcius, kelaparan sebab tak boleh nak buka puasa, last meal makan cuma dinner semalam, miss terus waktu maghrib....after a tiring day kat office memang lah cukup tak best. Bukan setakat tak best. Tapi memang palat lah.
The only consolation ialah waktu aku tengah dok tunggu taxi dalam gelap gelita depan office tu, boleh nampak bintang-bintang yang cukup cantik up above. Tak sangka Sydney yang crowded and banyak lampu tuh boleh nampak clear bintang-bintang kat langit.
Dalam ketensenan, dalam kelaparan, dalam kekalutan....ada keindahan. Wakakakaka.
Once in a while, I need to add brand new tapes as scratch into the tape library for a TSM client. Here's how :
tsm: SVRBKUP01>checkin libv TS3200_LIB1 025514L3 status=scratch checkl=no
ANR8432E CHECKIN LIBVOLUME: Volume 025514L3 is already present in library
TS3200_LIB1.
ANS8001I Return code 4.
This means that the tape is already checkin as scratch but we haven't label it. Someone on-site already put the tape into the library but haven't label it.
To resolve,
tsm: SVRBKUP01>checkout libv TS3200_LIB1 025514L3 checkl=no remove=bulk
ANS8003I Process number 942 started.
tsm: SVRBKUP01>label libv TS3200_LIB1 search=bulk checkin=scratch overwrite=yes
labels=barcode
ANS8003I Process number 943 started.
tsm: SVRBKUP01>q req
ANR8352I Requests outstanding:
ANR8373I 005: Fill the bulk entry/exit port of library TS3200_LIB1 with all LTO
volumes to be processed within 60 minute(s); issue 'REPLY' along with the
request ID when ready.
tsm: SVRBKUP01>reply 005
ANR8499I Command accepted.
What we did was to move the tape to bulk first. Bulk is basically the i/o station of the tape library. Then we ran the label libv command, as it would have labelled both.
By labelled, it labels the header of the tape by basing it on the barcode sticker.
From Melbourne to Sydney at 6:00am in the morning...
Almost there, 7:20am local time.
Sekarang ni kat bilik hotel in Melbourne. Hotel ni aku dah pernah lepak dulu, just around the corner from China Town. Comfy and senang nak dapat halal food. Tapi sat saja aku lepak this time sebab esok pagi-pagi aku dah nak kena catch 6am flight to Sydney.
Tadi meeting was fruitful. Fruitful lah kot. Ntah aku pun tak tau. Banyak expectation and banyak yang nak kena buat dalam 4-5 minggu akan datang ni.
Masa on the way from rumah nak ke airport, as usual aku akan sembang2 dengan driver taxi. Saja nak pass the time. Driver taxi tu still student, final year in accounting. Baru 2nd day bawa taxi, jadi jalan pun tak pandai...aku pula yang dok direct dia camna nak pi airport. Lupa nak cakap, every taxi kat sini memang equipped dengan GPS. Jadi kalau tak reti jalan pun, leh juga jadi driver taxi.
Then bila kat Melbourne, from airport nak ke hotel...driver taxi pun India mari juga. Student juga mamat ni. Ambik course hair-dressing. Katanya dia bukan nak jadi hair-dresser pun, tapi sebab tak dak course lain yang leh ambik..dia ambik juga. The reason is dia nak apply PR kat sini. Tak nak balik India lagi katanya. Dia baru 10 bulan kat sini. Bawa taxi 4 days a week, dari 5am to 5pm....and gajinya hanya $100 per day after tolak sewa taxi and duit minyak.
Macam-macam dia dok cerita. Katanya tempat dia susah, electric pun tak dak hari-hari. Kalau ada pun cuma 4 jam sehari. Banyak betul dia dok mengutuk India. Aku layan saja.
Bila sembang2 macam ni...buat aku terpikir. Betapa susahnya orang lain. Susah payah tinggal keluarga, tinggal segalanya nak dapat kehidupan yang lebih baik. Kerja almost 12 jam sehari dengan gaji yang ala kadar, lepas tuh nak kena study lagi pula for exams and all. So kita ni mostly sebenarnya very lucky. Tak perlu susah payah sangat, almost everything dah terletak depan mata.
Memang tak semua yang kita nak dalam life ni kita dapat. Tapi tak bermakna kita tak bernasib baik. We have just to count our blessings, strive harder and hope for the best.
Hari ni 13hb Ramadhan 1429H, ada majlis berbuka puasa kat Kent Senior High School anjuran orang-orang Melayu kat sini. Insya-Allah, pi kot petang nanti. Ada potluck dan juga gotong royong masak ramai-ramai. Acara start pukul 4:30pm, berbuka pukul 6:00pm and maybe terus jemaah Maghrib, Isyak and terawih sekali. Dengar kata ada juga acara bacaan Yassin dan tahlil. Mesti ramai berduyun-duyun orang yang mai. Sekarang ni kat sini, Melayu dah bersepah-sepah. Kalau jalan-jalan kat city, confirm selisih dengan Melayu lain. Kalau tak selisih dengan Melayu, mesti selisih dengan Mat Salleh. Hehehe.
Esok tengah hari pula aku dah berangkat ke Melbourne, ada meeting petang tu. Ada kaa meeting buat petang Ahad? Bongok betui. Tapi dah boss yang arrange, complain kat sini sajalah. Tidur satu malam kat Melbourne dan pagi Isnin dah naik flight at 6:00am to Sydney pula. Sampai Sydney should be around 7:30am camtu, aku rasa terus aku masuk ofis lah. Nak check-in hotel pagi-pagi buta sure tarak harapan. Kalau ikut schedule akan stay Sydney sampai Khamis, jadi tengah malam Jumaat tuh insya-Allah dah sampai balik ke sini lah kot.
Tuh saja nak update for now. Selamat.
This is the procedure to perform a failover test on Sun StorEdge 3510 array. The array has two controllers; primary and secondary; and are set in redundancy mode.
First check the actual status using sccli:sccli> show redundancy
Primary controller serial number: 8001100
Primary controller location: Lower
Redundancy mode: Active-Active
Redundancy status: Enabled
Secondary controller serial number: 8034367
Then, type "fail" command to fail the primary controller:sccli> fail primary
Are you sure? y
After that,the status should be like this:sccli> show redundancy
Primary controller serial number: 8034367
Primary controller location: Upper
Redundancy mode: Active-Active
Redundancy status: Failed
Secondary controller serial number: 8001100
At this point, the secondary controller should take over the disk arrays and the LUNs should still be accessible from the server to the disk arrays. Notice that the primary controller location is changed from Lower to Upper.
Next, type "unfail" command to bring back the primary controller:sccli> unfail
Are you sure? y
Check the status :sccli> show redundancy
Primary controller serial number: 8034367
Primary controller location: Upper
Redundancy mode: Active-Active
Redundancy status: Scanning
Secondary controller serial number: 0
sccli> show redundancy
Primary controller serial number: 8034367
Primary controller location: Upper
Redundancy mode: Active-Active
Redundancy status: Enabled
Secondary controller serial number: 8001100
Repeat the same "fail" and "unfail" commands for the secondary controller. All done for the failover test.
While I was doing the failover test, I had encountered some problems. The disk arrays was actually connected to two different Solaris servers. One is running Solaris 5.8 and the other is Solaris 5.9.
On Solaris 5.8, the failover test went well. The LUNs are all still visible and all filesystems are good. No issue here.
But on Solaris 5.9, several of the LUNs and filesystems were missing and Oracle crashed. I don't know why this is happening. At the moment, the workaround is to connect to remote system control (RSC) and reboot the server in reconfiguration mode.
First, I asked the DBA to make sure that none of the database are running. Next, I unmount all filesystems. Force unmount if necessary. Then at RSC, do :
rsc> sync
rsc> sync
rsc> reboot -- -r
The server should be back up in a few minutes and all LUNs are back online.
Later that night, I was informed by my colleague that Solaris 5.9 failed because of the SAN fcp driver issue. He said there is patch for it in Sunsolve and the patch number is 113040. Check if your patch is up-to-date :# showrev -p | grep 113040
From my mailbox...
One day a farmer's donkey fell down into a well. The animal cried piteously for hours as the farmer tried to figure out what to do. Finally, he decided the animal was old, and the well needed to be covered up anyway; it just wasn't worth it to retrieve the donkey. He invited all his neighbors to come over and help him. They all grabbed a shovel and began to shovel dirt into the well. At first, the donkey realized what was happening and cried horribly. Then, to everyone's amazement he quieted down.
A few shovel loads later, the farmer finally looked down the well. He was astonished at what he saw. With each shovel of dirt that hit his back, the donkey was doing something amazing. He would shake it off and take a step up. As the farmer's neighbors continued to shovel dirt on top of the animal, he would shake it off and take a step up. Pretty soon, everyone was amazed as the donkey stepped up over the edge of the well and happily trotted off!
Life is going to shovel dirt on you, all kinds of dirt. The trick to getting out of the well is to shake it off and take a step up. Each of our troubles is a stepping stone. We can get out of the deepest wells just by not stopping, never giving up! Shake it off and take a step up.
Remember the five simple rules to be happy:
1) Free your heart from hatred - Forgive.
2) Free your mind from worries - Most never happen.
3) Live simply and appreciate what you have.
4) Give more.
5) Expect less.
NOW...Enough of that crap....The donkey later came back, and bit the farmer who had tried to bury him. The gash from the bite got infected and the farmer eventually died in agony from septic shock.
MORAL FROM TODAY'S LESSON:
When you do something wrong, and try to cover your ass, it always comes back to bite you. Hehehe.
Sekarang ni dah masuk musim spring (toing toing!!). Jadi banyaklah pokok-pokok kayu yang jadi pokok-pokok bunga. :)
Pokok-pokok yang kecik belum start blooming lagi, tapi tak lama lagi lah tu. Nanti dah start blooming aku tangkap gambar...taruk banyak sikit bunga dalam blog ni kasik ceria.
Close up...
Gambar memang tak cun lah sebab camera aku ni al-maklum cap cipan saja, bukan macam setengah orang tuh pakai D70... :P
Sambil dok close-up nampak si rainbow lorikeet ni. Burung lorikeet memang suka gila lepak kat flame tree. Aku rasa kan memang ada madu kaa takpun buah kaa kat flame tree yang memang jadi makanan ruji si lorikeet. Kalau dekat WA, burung lorikeet memang bersepah-sepah banyak macam burung gagak bersepah-sepah banyak kat Msia.
Memang lah hampeh semalam, tak pasal-pasal aku kena menatang spyware AntiVirus XP 2008 ni. Dah siap pasang 2-3 anti-spyware, anti-virus, firewall dalam laptop ni pun leh kena getah juga. This is the first time kena since aku guna Windows XP sejak bertahun dulu. And yang first time ni pula yang paling tak sempoi. Antivirus XP 2008 ni memang kurang asam, aku try remove terus dia buat hal. Crashed terus office laptop aku ni. Mula-mula keluar blue screen of death as usual lepas tuh laptop ni asyik reboot tak henti-henti.
Nasib data aku semua aku letak kat partition lain. Dan nasib juga data aku tiap-tiap minggu rsync masuk external hard drive.
Memang tarak jalan lain, kena format and install balik semua. Nak format and install tak dak masalah tapi aku dok pikir abislah settings network connections aku semua. Dah laa ada beratus connection settings dalam tu...semuanya dial-up & VPN settings untuk client-client kat ofis.
Tapi sebab aku kan terer**....jadi jumpa la juga kat mana windows simpan settings tuu...
Kena cari file rasphone.pbk...
C:\Documents and Settings\All Users\Application Data\Microsoft\Network\Connections\Pbk\Rasphone.pbk
....copy file tuh dan restore balik dalam laptop yang serba baru.
Untuk putty settings pula, buka regedit cari keyword "SimonTatham" lepas tuh export key tu. Nak restore balik, just right-click file *.reg tu and choose Merge.
Dua benda yang paling penting dah settled. Sekarang ni tengah dok install balik semua benda dalam laptop. Banyak la pula applications aku nak kena download. Suci bersih berkilat seperti baru Windows ni. Hek.
** bukanlah terer pun tapi pakcik googlelah penyelamat...hehehe.
