Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: (ASCEND) Handy Hints For Upgrading Maxen
I wrote:
>> Despite the fact that Ascend's web-based "Upgrade Instructions" document
>> using a serial link as an upgrade tool, it simply "does not work" in the
>> words of Ascend tech support.
And Tim Basher replied:
>...The instuctions for the 7.0.0 release may be found on page 103...
Oh, come now. Ascend has a website, and puts up easy-to-find
web pages clearly labeled "Upgrade Instructions" but expects
customers to ignore them, and wade through over >>100 PAGES<<
of some OTHER document to find the CORRECT instructions for
upgrading?
What sort of practical joke is being played here?
>You can find the release notes in the doc/ sub-directory that can
>be found in the same directory you pick up the software.
...which itself was "hard to find", as pointed out in the
original e-mail.
>ftp://ftp.ascend.com/pub/Software-Releases/Max/Release-6.1.X/6.1.7/doc/r617.pdf
Oh, PDF. What a big help. Somewhat impossible to view
on a Palm Pilot (my new favorite hardware administration
terminal), as they are impossible to view with a mere web
browser.
Why even HAVE a website if the truth is buried in a series
of PDFs that can only be found by people with the navigation
skills of Magellan, do not show up in website searches, must
be completely downloaded to be read, can only be read with
a unique (non web) viewer, and are not even INDEXED on a web page?
>Note to Ascend -
>The standalone MAX Upgrade instructions are out of date, and do not even
>mention the Restricted and Extended loads.
Exactly. They do not mention ANY of the problems I encountered.
They are thus criminally misleading.
>But even these instructions make it clear that some types of MAX software
>loads require the use of TFTP.
But not which ones. Since I have never been a member
of the Psychic Friends Club, my guesses are less than
good enough to find the "correct" instructions amongst
all the misleading ones claiming to be correct instructions.
The "fact" of the serial port not working forces me to
question the integrity of the serial xmodem support.
Why CAN'T a "load" of any size be sent over the serial
port? I have never seen a file transfer problem with
any other hardware that supported xmodem, ymodem, kermit,
or ascii uploads, regardless of size.
>> The upgrade, once complete, erased all the options that had been
>> provisioned on the Max, even "Data Call".
>According to the people I have talked to at Ascend, the upgrade does not
>erase the options.
According to my own EYES, the upgrade DOES erase the
options. Please, your honor, strike the hearsay,
second-hand rumors, and apocryphal stories.
Before the upgrade, they were there. After, they were not.
Perhaps it was (again) alpha particles or an act of God,
but there seems to be a direct and proximate link of
causality too strong to ignore.
>The hash codes are stored in a different piece of hardware (along with
>the serial number). The problem is that the options were never enabled
>in the hash code in older units. The software did not check the hash
>code and just provided the feature. The newer software is now checking
>the hash codes - which works fine in newer units which have the correct
>hash code shipped from the factory - but breaks the operation
>of older units, and require the new hash code.
So, in short, the upgrade ERASED THE OPTIONS!! Why attempt
to claim that it did not?
No amount of techno-babble can explain away what happened.
It is not reasonable for this to happen, and in many
configurations, the "breaking" (or erasing) of the
hash-coded features would take the box offline from its
link, turning a remote "upgrade" into a self-inflicted
network failure.
Since one CANNOT use only the serial port or only the
network link to upgrade, this means that one must:
a) Go to the site where the Max is
b) Use tftp (over the network) to upgrade,
[or use the serial port for the "r*"
load, then use tftp for the "f*" load]
c) Then when the machine goes offline,
due to loss of mission-critical options,
HAND TYPE (or cut and paste using a
text file and a terminal emulator) the
hash codes into the box to restore the
options.
d) Then load a "saved copy" (brought on a laptop,
since the Max can be assumed to be offline
by this time) of the configuration
information back into the Max when the
loading of the hash codes erases the config.
What's wrong with this picture? Is this a 1990s approach
to upgrades? How do I justify a multi-million dollar
investment in a 24x7 NOC when long drives or plane flights
are required to upgrade a lousy (apparently, very lousy) RAS?
One cannot do a simple upgrade without BOTH a network link
and a laptop!!! First one link is required, then the other.
This is a new and unusual interpretation of the term
"administration" to which I have hitherto not been exposed.
If anyone would have told me that a hunk of hardware sold in
the 1990s would require me to budget for travel expenses
simply to do software upgrades, I would have laughed them out
of my office. It appears that I unwittingly approved the
purchase of a number of these boxes, and that these boxes are
clearly defective.
>Calling in advance to get the current hash codes is a good idea if you
>know to expect it. This has been greatly discussed on the Ascend Users
>mailing list.
But apparently, not yet FIXED by the vendor.
>Note to Ascend -
>The Release Notes do _not_ warn of this possible problem. This does
>cause problems for people and should be pointed out in advance
Yeah, that would be nice.
I would like to hear how Ascend thinks that any of this
could have been accomplished remotely. When they explain
the process, I will test it. If a remote box goes down,
as a result of the process they can pay someone to go onsite.
If it is not possible to administer these boxes remotely,
they are defective, and need to be replaced with something
that can be administered remotely. All of them.
>> After loading the hash codes sent by Ascend, the unit lost its
>> configuration, and had to be reloaded from a saved copy. Once
>> again, this could be a serious problem in a case where the Max
>> is a remote one, and connects back to the core location over a
>> T-1 plugged into the Max itself.
>
>The "update" command should not cause the unit to reset, so loading the
>hash codes should not casue the unit to lose its configuration.
Tim, it happened in front of my eyes. I have two more
Maxen of the same vintage with the same software release,
needing the same upgrade, so you can come and watch the
next one yourself if you'd like.
>However, with most software updates you need to follow the instructions
WHICH ONES????!!!!! :)
>carefully because you typically need to do an "fsave" and "nvramclear"
>to get the system to rebuild its configuration
Sorry, been there, done that, not a clueless idiot.
IT LOST THE CONFIG.
>Otherwise you run the risk of corrupting the information in NVRAM and
>losing your configuration.
Perhaps this is another "because it is more than a year
old, and was running the last known stable release of
code" sort of deals, where the "current product" does
not exhibit the failure, but "older" revs or hardware do.
I dunno. I don't care. I just wanted to be able to
FIND instructions, and be able to follow them to do
what was supposed to be a slightly boring, but trivial
task. I did not want a Cecil B. DeMille production.
(What I got was a Stephen King novel!)
Can ANYONE document or suggest a method of upgrading
these things that will not force a site visit, or
must we now discuss the "gas mileage" of a Max in
terms of thousands of miles driven per Max per year?
If so, we must start cross-posting to rec.cars.volvo!
"...Lead us not into temptation, but deliver us some e-mail. Amen"
james fischer jfischer@supercollider.com
++ Ascend Users Mailing List ++
To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd: <http://www.nealis.net/ascend/faq>