Disk failure
One thing to be done today was another attempt to repair the Windows installation, but there must have been something very wrong, because it seemed to take hours. So I aborted the installation, accepting I would have to reinstall all software.
Starting with Office 2007 – but that failed for some reason.
One other thing I would do is install more memory. The shop offered me a 2Gb DDR-2 DIMM, but the system cannot handle this type of DIMMs – the notch is on a different location so the DIMM didn’t fit.
When I next tried to reboot the system, the first disk was gone in BIOS….
It appears to spin so that is not an issue. I checked power and SAS cable, switched with the other drive, but that didn’t work either. So probably there is something wrong with the electronics on the disk….
A colleague has the ability to try to restore data from broken disks, in some extent. And since there is some data on that disk that is not copied to another location, it’s worth to try to get it off the drive. Otherwise, it’s lost. Pity, but not fatal.
23-Jan-2010
New kids on the block
I got my niece’s old ACER laptop, it has a non-working TFT screen and a CD drive that isn’t reckognized by Windows (it seems) but otherwise running fine. That one (nicknamed “Melvin” – after her cat) will be used in testing OpenVMS 8.4 when running PersonalAlpha (and do nothing else).
PersonalAlpha on my company laptop fails to connect over the Loopback interface, but I think that’s due to the fact that I had to disable some interfaces for the VPN connection …
Speaking of PA: It isn’t the only free Alpha-emulator any more. Camiel Ver der Hoeven started working on a ES40 emulator a few years ago, and Bruce Claremont of Migration Specialties joined in to complete the CPU emulation. There will be a commercial version some time this year, but the free version is available in BETA test; the first versions were Windows 64 only, but since last week, a 32-bit version is added.
This has now been installed on my normal workstation – to test it, and to add into the 8.4 fieldtest as well. I started installing it:
It got as far as obtaining the initialisation data for the system disk, but than run into a runtime error… Well, it’s still beta, it’s allowed in that phase :). Ok, contact Bruce and retry.
Next week, I;’ll spend some time in setting up the Cluster-over-IP environment. The links have been set up.
06-Jan-2010
New router activated
Setup of the new router being completed – at least, as far as I could do without digging the documentation – I tested the wireless connection to make the last corrections: Change of the interla IP address so it matches the current router. Just in case it would be better to restore the Linksys router, when the Draytek didn’t work as expected. Next, I switched all internal and external connections, throw the power switch of the new router – and waited a while. It takes some time to establish all connections with all logging enabled: getting the connection data (DHCP), connect to the DNS servers and negotiate; setting up internal connections (DNS, for instance), AND handle requests.
First I tried to access the webs using the external names, but it took a few minutes before it succeeded.
Accessing the World Wide Web proved to be easier – I could access extrenal webs before I could access my own. From an external email account, I could send a message – and it arrived in either quarantine (by the spam filter) of in my mailserver. So mail (quite important) is no problem either.
Next, I tried to access the login host of my ISP, but that failed competely. It turned out I would have to set up something specific, after that, TELNETting to my ISP was no problem. From there, I could use Telnet and SSH to access Diana – after I figured out I have made a minor mistake in the port definition.
FTP however is another matter. Both anonymous and non-anonymous FTP fail when data is to be returned – I can make a connection but DIR fails. Setting mode to passive makes access possible. I’ve seen that behaviour before but I cannot recall what’s the cause. It must be some setting in the router’s firewall or NAT definition, that’s for sure. Time to RTFM 🙂
The VPN setting has been prepared but to test it, I’ll need to be outside. Something to test tomorrow.
ISDN to return
This modem can be connected to ISDN directly, and it offers the ability for Voice-over-IP, and ISDN backup. But for that, the ISDN cable – currently used to transport the ADSL signal to my modem – must be restored to it’s original function. Question now is how to get ADSL to my modem? I could of course move the modem near the splitter, but than I’ll have create a connection by CAT-5 cable to the bottom of the house. Something I’ve been considering for quite some time. Perhaps, this is the moment to do it.
Another possibility is using a second IDSL-cable to tarnsport the ADSL signal – that may well be a better solution, there is very little room left…
04-Jan-2009
New router
Just before Christmas, I ordered a new router, and after I came home on the last working day last year (Wednesday), I got the message the router was ready to be collected; If I only had that message BEFORE leaving the office – the supplier is more or less on my way home…
Anyway, this afternoon I got my new router, but before setting it up in the network, I did some setup, mainly the ports to be opened (all passing to Diana), some external ports mapped to standard ports (also on Diana) and some other additional setup: VPN, wireless, and matters on the internal network like the internal address. Of course, it’s a new address to be able to do the setup from a system in the network, and when replacing the Linksys router, the DHCP settings in Diana will need to be updated to refelct this change: It’s the default gateway and DNS resolver….. Or I’ll change the address when the router is separated from the network, before installing it. No chnages elsewhere in the network.
I’ll still have to do a bit of testing, but as far as I can determine, it looks fine. It works as an access point, anyway.
Replacement of the external gateway means I’ll be offline for a while – a short while, I hope. Somewhere next weekend, propably.
15-Oct-2009
New kid on the block
The new Personal Workstation I obtained a few weeks ago didn’t boot properly. Whatever I tried, it kept booting into ARC – the (graphical) NT console. First of all, I assured I had the latest software – had to get it in ROM using the FAT formatted floppy disk, and choose to update the firmware. That didn’t show any errors so it may be assummed it worked.
Next came the question how to force it into SRM – the VMS /Unix console. It took some digging in the docs but I finally found it: Boot into ARC, hit F2 to enter setup, tab to CROM settings and hit F6 to get into the advanced mode – where the right console can be choosen, hit F10 twice, then Enter, and last Escape – and cycle power. (Is this the famous “intuitive” interface?)
But on reboot, ARC re-appeared and much to my surprise, the console was reset to be “NT (AlphaBIOS)”.
More surprising: ARC mentioned the machine to be a PWS 500a – where the front bezel clearly stated it to be a PWS 500au – the first will only run NT, and for VMS (and Unix) you need an au-type! But given the mentioning on the exteriour and the presence of the board containing the 2 Mb cache memory, it indeed looked like a au – type.
👿 What was going on here?
I put the question on the OpenVMS ITRC forum, and I got the answer within an hour: The battery might be dead. On second thought, that was feasable since ARC constantly mentioned the time was invalid and set to factory default (1-Jan-1995). Other settings were fine, though, and changes were persistent over boot. After replacing the battery (a cost of 4 eruo’s…) I could finally boot into SRM ( 😀 Hurray !!!!) and prepare for installation of OpenVMS. So the BIOS data may be split into flash-memory (keeping settings without power) and cells requireing battery backup. It’s the only reason I can think of, for this weird behavior
Next I installed OpenVMS 7.3-2 on the box, for the time being anyway, mainly to see what’s in the box: 256 Mb of memory (I know that already) and one (!) 4.1 Gb disk….Well, that will do for now.
Tomorrow, or this weekend, I’ll try to get it booted from the shared SCSI – the second SCSI controller is a KZPBA, though from a different type that is located in Diana: KZPBA-CX in stead of KZPBA-CY. It has already been set up to host on LUN 5 (the HSZ50 runs on LUN 7, the Diana host is set on 6, so 5 would be the right choice here.
The machine will be known as “Daphne” – and if possible, it will run clustered with Diana.