Discussion:
hoffman labs?
(too old to reply)
Neil Rieck
2017-02-27 16:52:52 UTC
Permalink
Raw Message
While this URL is up:
http://hoffmanlabs.com/home.html
lab links like this are not:
http://labs.hoffmanlabs.com/node/16

Hopefully Steve will see this post.

Neil Rieck
Waterloo, Ontario, Canada.
http://bellics.com/ovms/
Stephen Hoffman
2017-02-27 17:19:32 UTC
Permalink
Raw Message
Post by Neil Rieck
http://hoffmanlabs.com/home.html
http://labs.hoffmanlabs.com/node/16
Separate software, separate locations, separate servers.
Post by Neil Rieck
Hopefully Steve will see this post.
Yes, the labs server has been reported as offline. It was pulled down
and access blocked at the firewall for some expected hardware upgrades
— rolled out a new deployment and new hardware, and the server is up
and running internally — and then somebody dropped a nasty RCE in some
of the software, and dealing with the ensuing dependency chain of
software upgrades hasn't been a priority. Particularly given it's a
non-revenue-generating web site. Having the server offline has
reduced the effort dealing with the folks trying to break into the
site, the folks trying to spam the site, watching for critical patches,
and the rest of the "fun" inevitably involved with that particular
community project.
--
Pure Personal Opinion | HoffmanLabs LLC
Craig A. Berry
2017-02-28 02:48:38 UTC
Permalink
Raw Message
Having the server offline has reduced
the effort dealing with the folks trying to break into the site, the
folks trying to spam the site, watching for critical patches, and the
rest of the "fun" inevitably involved with that particular community
project.
Yes, but 90% of the useful information about VMS available online has
been unavailable for a few weeks. Which is a bummer.
Phillip Helbig (undress to reply)
2017-02-28 19:51:34 UTC
Permalink
Raw Message
Post by Craig A. Berry
Having the server offline has reduced
the effort dealing with the folks trying to break into the site, the
folks trying to spam the site, watching for critical patches, and the
rest of the "fun" inevitably involved with that particular community
project.
Yes, but 90% of the useful information about VMS available online has
been unavailable for a few weeks. Which is a bummer.
Should have been running on a multi-site VMS cluster. :-|
Stephen Hoffman
2017-02-28 22:05:35 UTC
Permalink
Raw Message
Post by Phillip Helbig (undress to reply)
Post by Craig A. Berry
Having the server offline has reduced the effort dealing with the folks
trying to break into the site, the folks trying to spam the site,
watching for critical patches, and the rest of the "fun" inevitably
involved with that particular community project.
Yes, but 90% of the useful information about VMS available online has
been unavailable for a few weeks. Which is a bummer.
Should have been running on a multi-site VMS cluster. :-|
Could you elaborate on that?
--
Pure Personal Opinion | HoffmanLabs LLC
Phillip Helbig (undress to reply)
2017-03-01 05:44:18 UTC
Permalink
Raw Message
Post by Stephen Hoffman
Post by Phillip Helbig (undress to reply)
Should have been running on a multi-site VMS cluster. :-|
Could you elaborate on that?
Just a joke. Note the smiley.
att3b2
2017-03-01 12:06:05 UTC
Permalink
Raw Message
Post by Phillip Helbig (undress to reply)
Post by Stephen Hoffman
Post by Phillip Helbig (undress to reply)
Should have been running on a multi-site VMS cluster. :-|
Could you elaborate on that?
Just a joke. Note the smiley.
Maybe his website was in the Amazon Cloud :)
Stephen Hoffman
2017-03-01 15:46:18 UTC
Permalink
Raw Message
Post by Phillip Helbig (undress to reply)
Post by Stephen Hoffman
Post by Phillip Helbig (undress to reply)
Should have been running on a multi-site VMS cluster. :-|
Could you elaborate on that?
Just a joke. Note the smiley.
Ah, so the joke was spending what would probably be well past a hundred
thousand dollars to host what is a free web site, buying hardware and
software that does not and will not address the bottleneck, and that
would serve to compound the maintenance and upkeep effort and the
ongoing costs? Well, there could also have been a suggestion to port
the entire content management system over to Fortran. That would have
been... hilarious.

No, I'm not going to implement a multi-site OpenVMS DT cluster for the
purpose of hosting a free web site. The replacement hardware and
software that's already in use — well past OpenVMS in capabilities,
much simpler to configure and manage and maintain, and better suited
for this particular purpose — cost less than a one-core OpenVMS I64
license, too. If I wanted to spend rather more than what I already
have, I'd hire the necessary staff or outsource it, and that'd still be
cheaper than a DT configuration. If I did need a more available
configuration — still don't need DT — there are other better and
cheaper solutions for hosting a web site, too.

As for the content management system and the rest of the dependencies
that are blocking exposing that new server, that's a project for
another day. But that effort and resolving those dependencies would
arise on OpenVMS, with other issues added due to the somewhat
problematic web support present in OpenVMS.

BTW, the conventional interpretation for that particular :-| emoticon
(😐) doesn't align with "it's a joke", either. Hence the confusion.
--
Pure Personal Opinion | HoffmanLabs LLC
Simon Clubley
2017-03-01 18:37:41 UTC
Permalink
Raw Message
Post by Stephen Hoffman
BTW, the conventional interpretation for that particular :-| emoticon
(?) doesn't align with "it's a joke", either. Hence the confusion.
You weren't the only one to notice that. I also thought Phillip was
being serious.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world
David Froble
2017-02-28 22:16:32 UTC
Permalink
Raw Message
Post by Phillip Helbig (undress to reply)
Post by Craig A. Berry
Having the server offline has reduced
the effort dealing with the folks trying to break into the site, the
folks trying to spam the site, watching for critical patches, and the
rest of the "fun" inevitably involved with that particular community
project.
Yes, but 90% of the useful information about VMS available online has
been unavailable for a few weeks. Which is a bummer.
Should have been running on a multi-site VMS cluster. :-|
What possible help would that have been?

I do believe that Steve had other concerns.
s***@ieee.org
2017-03-01 01:36:49 UTC
Permalink
Raw Message
Post by Stephen Hoffman
Post by Neil Rieck
http://hoffmanlabs.com/home.html
http://labs.hoffmanlabs.com/node/16
Separate software, separate locations, separate servers.
Post by Neil Rieck
Hopefully Steve will see this post.
Yes, the labs server has been reported as offline. It was pulled down
and access blocked at the firewall for some expected hardware upgrades
— rolled out a new deployment and new hardware, and the server is up
and running internally — and then somebody dropped a nasty RCE in some
of the software, and dealing with the ensuing dependency chain of
software upgrades hasn't been a priority. Particularly given it's a
non-revenue-generating web site. Having the server offline has
reduced the effort dealing with the folks trying to break into the
site, the folks trying to spam the site, watching for critical patches,
and the rest of the "fun" inevitably involved with that particular
community project.
--
Pure Personal Opinion | HoffmanLabs LLC
The "LABS" are a wonderful resource. I find that they are often the best resource when I'm looking for assistance and clarification. Hoff's write ups are often my first choice for learning/relearning how to do something in the OpenVMS or Alpha/Itanium hardware worlds. The posts to the write ups by the community are often clarifying and almost always helpful. Their lack of availability has been a wake up call to me. I hadn't realized how much I used them until they weren't readily available.

So, thanks for your involvement in this "community project".
Neil Rieck
2017-03-01 12:28:47 UTC
Permalink
Raw Message
Post by s***@ieee.org
Post by Stephen Hoffman
Post by Neil Rieck
http://hoffmanlabs.com/home.html
http://labs.hoffmanlabs.com/node/16
Separate software, separate locations, separate servers.
Post by Neil Rieck
Hopefully Steve will see this post.
Yes, the labs server has been reported as offline. It was pulled down
and access blocked at the firewall for some expected hardware upgrades
— rolled out a new deployment and new hardware, and the server is up
and running internally — and then somebody dropped a nasty RCE in some
of the software, and dealing with the ensuing dependency chain of
software upgrades hasn't been a priority. Particularly given it's a
non-revenue-generating web site. Having the server offline has
reduced the effort dealing with the folks trying to break into the
site, the folks trying to spam the site, watching for critical patches,
and the rest of the "fun" inevitably involved with that particular
community project.
--
Pure Personal Opinion | HoffmanLabs LLC
The "LABS" are a wonderful resource. I find that they are often the best resource when I'm looking for assistance and clarification. Hoff's write ups are often my first choice for learning/relearning how to do something in the OpenVMS or Alpha/Itanium hardware worlds. The posts to the write ups by the community are often clarifying and almost always helpful. Their lack of availability has been a wake up call to me. I hadn't realized how much I used them until they weren't readily available.
So, thanks for your involvement in this "community project".
I agree. When HP was splitting their site into HP and HPE in 2015, most of their stuff was offline or missing (lots of stuff has vanished forever) while resources at Hoff's site saved my bacon.

###

On a related note, my OpenVMS resources site here:

http://www3.sympatico.ca/n.rieck/links/openvms_resources.html

...is a rinky-dink offering compared to Hoff's site. For example, I don't employ a CMS (content management system) or allow user feedback. Anyway, I plan to retire in 2020 so have been moving my OpenVMS stuff to here:

http://bellics.com/ovms/

Since I started this I can't believe the number of attacks that have been launched against this site. For example, I can't believe the amount of SPAM sent here. Last October someone injected 40,000 emails which this Itanium was trying to forward elsewhere (I thought I had already dealt with the SPAM holes). After I fixed that problem I noticed a continual probing via Telnet, FTP, and SSH. There were only two active accounts on this system, and the passwords where greater than 20 characters but eventually someone is going to get in (STUXNET?) so I closed every service port except 25, 80, 443, and 22 (SSH). I then restricted SSH to only accepting connects from a small subset of domains and even then, only allow one user (with a 20-character password) through.

I only mention this because Hoff's site must have been seeing attacks like this in spades, and it requires a lot of work to keep up with the world-wide-web of hackers. Not sure if Hoff was getting remuneration from HP for his efforts, but with the announcement by HPE to no longer offer standard patch support for "OpenVMS-8.4 on Alpha" starting Jan-1-2017, I think this indicates that HPE is moving on.

http://h41379.www4.hpe.com/openvms/openvms_supportchart.html

Neil Rieck
Waterloo, Ontario, Canada.
Stephen Hoffman
2017-03-01 18:11:19 UTC
Permalink
Raw Message
Post by Neil Rieck
...is a rinky-dink offering compared to Hoff's site. For example, I
don't employ a CMS (content management system) or allow user feedback.
Once site logins and user registrations are available and the related
registration-traces then become visible to the search engines, the
keeping-a-lid-on-it effort escalates. Don't Read The Comments doesn't
begin to cover what can happen here, though I've been able to keep most
of that off the site. More than a few entities actively search the
web for open-registration web sites, and they have and use automated
registration and spamming tools.
Post by Neil Rieck
Since I started this I can't believe the number of attacks that have
been launched against this site. For example, I can't believe the
amount of SPAM sent here.
Spam to the contact addresses, spam to the return addresses for the
registration mail, spam through the site contact links, spam runs to
other folks using those addresses as return addresses for the spam, and
more than a few attempts to post spam as comments, among other bits of
"fun". Based on metadata, the vast majority of the site registrations
are from one-shot visits or from bots; from automated tools.
Post by Neil Rieck
I only mention this because Hoff's site must have been seeing attacks
like this in spades,
Anybody running a site is, and it's only getting worse.
Post by Neil Rieck
and it requires a lot of work to keep up with the world-wide-web of
hackers. Not sure if Hoff was getting remuneration from HP for his
efforts,
The HoffmanLabs web site is solely funded by HoffmanLabs. While that
might obviously change in the future, there are no plans to do so.
While there've been suggestions around setting up a donation link, the
current "vacation" is rather less about the funding than about juggling
the time investment involved. Dealing with the web site, as well as
creating new and updating content and what few comments are queued.
Slogging through debugging and updates in general and specifically for
that pesky RCE in an underlying component, as is the present case.
Sometimes slogging on very short notice, and the notices are getting
shorter. There are other and higher-priority projects currently
seeking that same time. If this were exclusively about payment, I'd
just pay somebody to deal with this stuff. And I'm not looking to do
that, as it's a free web site.
Post by Neil Rieck
but with the announcement by HPE to no longer offer standard patch
support for "OpenVMS-8.4 on Alpha" starting Jan-1-2017, I think this
indicates that HPE is moving on.
HPE have been telling folks that HPE support for OpenVMS is ending, and
for some years. That's been the purpose of the roadmap.
Presentations on migrations became quite common there for a while, too.
It would not surprise me to see HPE entirely end their OpenVMS
efforts including their VSI OpenVMS reseller role once HPE exits the
Itanium business, as the end of the HPE OpenVMS roadmap is reached,
too. VSI will then be the sole outpost of OpenVMS, and for as long as
they remain sufficiently solvent.
--
Pure Personal Opinion | HoffmanLabs LLC
j***@gmail.com
2017-03-02 12:55:20 UTC
Permalink
Raw Message
Post by Stephen Hoffman
Once site logins and user registrations are available and the related
registration-traces then become visible to the search engines, the
keeping-a-lid-on-it effort escalates.
One idea to consider, assuming that Hoff would approve of this strategy,
would be to harvest some of the content and "re-deploy" it to websites
such as stackoverflow.

Some of his articles could probably be repurposed into "questions" and
then answers provides. Others would probably fit well into the documentation
system.

Stackoverflow and the many stackexchange websites are a terrific
resource. It seems like much of Hoff's content could live on there. This
would free Hoff of the operational obligation of maintaining a website
in the modern world and would turn it into a community effort.

EJ
Stephen Hoffman
2017-03-02 16:25:01 UTC
Permalink
Raw Message
Post by j***@gmail.com
Post by Stephen Hoffman
Once site logins and user registrations are available and the related
registration-traces then become visible to the search engines, the
keeping-a-lid-on-it effort escalates.
One idea to consider, assuming that Hoff would approve of this
strategy, would be to harvest some of the content and "re-deploy" it to
websites such as stackoverflow.
That's not happening.
--
Pure Personal Opinion | HoffmanLabs LLC
Neil Rieck
2017-03-08 15:53:49 UTC
Permalink
Raw Message
Post by j***@gmail.com
Post by Stephen Hoffman
Once site logins and user registrations are available and the related
registration-traces then become visible to the search engines, the
keeping-a-lid-on-it effort escalates.
One idea to consider, assuming that Hoff would approve of this strategy,
would be to harvest some of the content and "re-deploy" it to websites
such as stackoverflow.
Some of his articles could probably be repurposed into "questions" and
then answers provides. Others would probably fit well into the documentation
system.
Stackoverflow and the many stackexchange websites are a terrific
resource. It seems like much of Hoff's content could live on there. This
would free Hoff of the operational obligation of maintaining a website
in the modern world and would turn it into a community effort.
EJ
I'm assuming he has it all backed up. There is a lot of open source server software out there ( https://en.wikipedia.org/wiki/List_of_content_management_systems ) so I wonder if he could recover the data to a different platform. If the previous machine was an Itanium then he might need to move to an x86-XEON (probably an HP ProLiant)

p.s. I had to do something similar recently. I grabbed an old DL-385-G7 then loaded CentOS-7 just to test some open source software.

http://www3.sympatico.ca/n.rieck/docs/openvms_notes_linux.html

Neil Rieck
Waterloo, Ontario, Canada.
http://www3.sympatico.ca/n.rieck/
Arne Vajhøj
2017-03-09 03:08:21 UTC
Permalink
Raw Message
Post by Neil Rieck
I'm assuming he has it all backed up. There is a lot of open source
server software out there (
https://en.wikipedia.org/wiki/List_of_content_management_systems ) so
I wonder if he could recover the data to a different platform. If the
previous machine was an Itanium then he might need to move to an
x86-XEON (probably an HP ProLiant)
There are hundreds and hundreds of open source CMS solutions
(most PHP, some Java, some .NET, some Python).

A Linux box with a CMS is so easy.

Some of the stuff should run on VMS as well if so desired.

:-)

A CMS will make it a lot easier to maintain pages.

But keeping the server uptodate and secure still requires work.

Arne
Stephen Hoffman
2017-03-09 13:42:11 UTC
Permalink
Raw Message
I'm assuming he has it all backed up...
Yes, there are backups. Nightly database backups, and periodic
backups of the web site configuration, and separate system backups.
Etc.

The web server and content management system is also presently up and
running, and is internally accessible, and is serving the HL site
contents. From newly-acquired hardware, running current system
software and the same and problematic content management system version.

External access to the web server and the current content has been
firewalled, pending updates and resolution of the RCE within the
content management system and particularly within a dependency that's
in use locally.

I'm not inclined to port the contents to another platform nor deal with
all that entails, as an upgrade of the current system is almost
certainly going to be easier. Nor am I inclined to install a Xeon
server, as a newly-acquired Core i5 is more than sufficient for what is
a free web site.

There are simply some other priorities right now.

The Labs server will be back online when it's upgraded and able to be
brought back online.
--
Pure Personal Opinion | HoffmanLabs LLC
j***@yahoo.co.uk
2017-02-28 08:18:12 UTC
Permalink
Raw Message
Post by Neil Rieck
http://hoffmanlabs.com/home.html
http://labs.hoffmanlabs.com/node/16
Hopefully Steve will see this post.
Neil Rieck
Waterloo, Ontario, Canada.
http://bellics.com/ovms/
OK, so we now know it's currently unavailable and we know
that the site owner is aware.

If it's the former content that's of interest, rather than
the notification of the unavailability...

What approach would readers normally use when any website no
longer responds as expected, or a previously useful web page
vanishes for some reason?

I know what I'd usually start with. And for the particular
page in question, it currently does what I'd hope it would
do. Quite simply too, once you work it out.

The approach is not VMS specific nor is it site specific,
but it does have some administrative and technical limits
which occasionally restrict its usefulness.

E.g. the approach does require the current site owner to not
object e.g. much content from the DEC, CPQ, and HP websites
is no longer accessible via this mechanism because the
current site owner(s) representatives have requested their
content (which had previously been available) be removed
from this mechanism. There are doubtless other reasons why
this approach may fail in some cases, but it's often worth
a try.

Hint: Wayback machine at archive.org.

I tried this a few minutes ago on the page in question,
leading to (e.g.) this version captured in 2016.

http://web.archive.org/web/20160223111241/http://labs.hoffmanlabs.com/node/16

Yes that link is meant to look like two URLs concatenated.
Here's the saved content it found for me just now:


OpenVMS Custom Services and Support from HoffmanLabs
Submitted by Hoff on December 30, 2006 - 05:12

OpenVMS OpenVMS Alpha OpenVMS I64 OpenVMS VAX Services

Do you have a conundrum?

HoffmanLabs LLC specializes in HP OpenVMS production systems, enterprise-class hardware and software, clustering and uptime, as well as Apple Mac OS X and Mac OS X Server configurations.

[continues]
Loading...