Re: [cobalt-users] installed upgrades, Qube2 sendmail has stoppedworking

[Malcolm McLeary <mmcleary@xxxxxxx>]

Hi Lawrence,

on 3/8/01 3:57 PM, Lawrence Lee at tomalak@xxxxxxxxx wrote:

> Could you give me some more background on what kind of things to watch
> out for when patching a Qube2? I have a machine sitting on a server
> farm, so I'm a bit nervous about patching it and having it die on me.

That is a tough one.

I've had my fair share of disasters with the original Qube2700WG and Qube2.
To date I've not had problems with the Qube3 but that may be because I'm
actually prepared for possible disaster.

On the Qube2700WG one of the major OS updates completely hosed the GUI on my
Qube.  To make matters worse it was one of the "originals" which didn't (and
still doesn't) have the firmware to netboot ... hence an OS Restore CD is
useless.  Fortunately it was only the GUI ... all the data was intact so
after running the builtin system reset process I was able to get it going
again.

As for the Qube2, there was an early OS Update which stopped Appleshare
services.  Then there was another which killed the standard web server, but
left the admin web server running ... it you didn't know the url to get
straight to the admin GUI it looked bad.  More recently I updated a Qube2
and once again lost Appleshare services ... its still down and Cobalt/Sun
are escalating it.

Best advice;

1. Don't be the first to install an update or a patch.

2. Only ever install one at a time and verify all is well before doing the
next.  

3. If at all possible backup your system before doing an update.

4. Be there ... I hate remote updates ... if something bad is going to
happen, it will happen when access to the box is inconvenient.

5. If possible, don't install updates on production systems until you have
tried it on a development unit, but how many Qube owners have a "spare".

With the Qube3 updates seem to be of a much higher quality, but I still try
them on a development unit before doing it on a production unit.  I like the
Pro Ed because I can replace the second disk with a spare and let the RAID
software sync the disks before doing the update ... if the update goes bad,
I replace Disk A with the Disk B I swapped out before doing the update and
I've effectively restored the system to where it was before the update.

Cheers,  Malcolm

-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --

                       Information Alchemy Pty Ltd
                             ACN 089 239 305
                           Canberra, Australia

Malcolm McLeary                                Mobile:     0412 636 086
Managing Director                              Email:  mmcleary@xxxxxxx

     This message was sent using Outlook Express 5.0 for Macintosh.