Friday Oct 10, 2008

Login to sunsolve just once a day

Go on. 7 ½ hours before you need to use SunSolve login and then just leave that tab alone until you need it. Why? Because you can!

As promised the horribly short idle time out has been increased from 30 minutes to 8 hours and the session time from 2 hours to 24.

Also I as have just been reminded it effects blogs.sun.com too, sweet.

Tuesday Sep 30, 2008

Sunsolve et al Session timeouts increasing.

The hot news around here is that the session timeouts for Sunsolve and the other tools that use the authentication system on sun.com are going to be increased to something approaching reasonable timeouts. The current 30 minute idle and 2 hour session timeout will be increased to 8 hours idle and 24 hours for the session. Not quite the 14 days and 90 days I would have but none the less a welcome step in the right direction.

If all goes well the change should happen on October 9th. I wish it was sooner but none the less the prospect is exciting enough for me to pre-announce it here, not that anyone will read it!

A big thank you to those who are making it happen.

Saturday Sep 20, 2008

Office Advert

There is an amazing Advert running on local radio at the moment. The premise is that you don't need to waste money on expensive brand name trainers but to get the best out or education then you have to have Microsoft Office 2008. The irony of the add is that you would be wasting your money not on brand name trainers but instead on a“brand name” office suite.

Of course for infinitely less, yes free, you can have OpenOffice.org.

I know I've been here before but it is worth repeating.

Friday May 02, 2008

SunSolve Search

I've added a completely unsupported Firefox Search engine for SunSolve to this blog. It you go to the search box on Firefox while viewing this page then it will give you the option to add the search engine. For the search engine to work you have to be a registered SunSolve user and if/when SunSolve changes it could stop working.

I'm going to log an RFE for SunSolve that they include this in SunSolve itself at which point I'll delete this one.

If anyone has a better icon to use then I would be happy to include it.

Monday Mar 31, 2008

Sun Ray @ home downside and upside

The downside is that one day you will ride to work only to find you have an empty pass holder in you pocket.

The upside is that you can check and see that you have left your pass in your Sun Ray @ home:

: enoexec.eu FSS 1 $;  utwho -c | grep 'JavaBadge.\*cg13442'

 71.0 JavaBadgeNP.4090009c2311b2071914 cg13442  129.150.116.130 P8.00144f7dc334

: enoexec.eu FSS 2 $; 

So I know it is not lost. Now the problem is that it contains all the cash for lunch in a “cashless office”.

Friday Feb 15, 2008

20 years.....

Twenty reasons, in no particular order, why working at Sun has been and still is a blast.1

  1. NFS. No really it is better then rcp. Which is saying something as rcp is much better than uucp and uucp seemed much better than sneakernet.

  2. YP NIS. Actually I hate NIS but the simplicity of it for solving a problem was in it's time fantastic. That it is still in use is amazing or depressing depending on the time of day2.

  3. Mailtool. A graphical tool for reading your email. It made giving up on getting a vt220 worthwhile.

  4. The Network is the Computer”. Now that is a slogan I could then and still do really get.

  5. The Systems3:

    1. 3/160. I blagged one of these as my workstation as it had 12 meg of memory. It was being returned as upgrade residue I only got rid of it when I was offered a:

    2. 4/110. A workstation capable of running NeWS, well if you had the floating point unit. My one was a lemon, if ever any one heavy walked past my desk it would reset. However when it was running it was fast!

    3. SparcStation 1. Audio as standard on a workstation. How cool was that.

    4. The 386i. Look what happens when you take a slightly different approach to system management. Yes all your hosts are called “oak”.

    5. The SPARCstation 10. Four CPUs on the desktop, fab, ISDN on the back, not so fab.

    6. The SPARCcentre 2000. Have some more CPUs. You know you want them. (Then for laughs naming the upgrade the 2000E, dooming our lab system to be known as Cortina.)

    7. The E10000, have some more CPUs and DR.

    8. The T2000. Go on put all those CPUs on one chip.

  6. The SPARC Storage array. Put all those disks in one box. Attach two fibres.

  7. NeWS. A window system that made a virtue of being able to program in PostScript, allowing those of us who were intimate with the printers to be the experts!

  8. TFS. The translucent file system. It had more subtleties than just being translucent but you could build some brilliant technologies4 on top of it. I wanted to smoke some of the stuff whoever thought it up was smoking.

  9. The automounter. Even when it used symbolic links and /tmp_mnt to mount things it was cool. Since it has had autofs it is positively sub zero, don't start me on the things you can do with executable maps.

  10. Threads in user land. Not those light weight processes we had in 4.x real fully fledged threading in a single process. Now my mail search program could consume the whole SC2000, briefly, very briefly.

  11. OpenLook. You gotta love those push pins.

  12. Solaris5

    1. Solaris 2.3. An OS capable of getting the most out of the SC2000.

    2. Solaris 7. How much address space would you like sir?

  13. Email attachments. Obviously we were completely happy using uuencode and uudecode the arguments just slipped from your fingers but attachments seem to have caught on anyway.

  14. Sun Ray. I wonder how people manage home computers without Sun Ray they are quite useful in the office too.

  15. Oak Java. The only reason anyone I know outside of IT has heard of Sun before I tell them.

  16. Blogs.sun.com. You allow, no, encourage your employees to write about “anything”. It's called trusting your staff. All companies should do it.

  17. OpenSolaris. Put the source out in the open and see what happens.

  18. Dtrace. Only because I can and do use it in an executable automount map.

  19. ZFS. At last we can hold our heads up and say that the file system is going to have the contract with the user that the users always thought that they had. Ie you will get back the same data as you wrote or an error.

  20. Putting up with me for 20 years!

I've not included the people which are the real stars of which there have been and are many singling them out somehow seems wrong.


1I don't claim all of these were invented by Sun although many were.

2A bit like working at the same company for 20 years.

3Yes I cheated putting all the systems in one item.

4NSE does not count as a “brilliant technology”. It could have been but was not. I know I used to support it.

5Yes another cheat.

Thursday Jan 17, 2008

Opposite's in Service

When I read the announcement about Sun's acquisition of MySQL on Jonathan's blog I too was drawn to the fact that for every employee of Sun Service there is an equal and opposite employee. Check out the picture:




I'm left wondering what my opposite would be like. Clearly they would not ride a bike. I hesitate to suggest asking what other characteristics they would have but no one reads this anyway.

Monday Nov 19, 2007

Sun internal uk-solaris meeting

On Wednesday we will have the first of what I hope will be quarterly UK Solaris meetings. Years ago these meetings were held on a semi-regular basis and allowed engineers in Sun who were in the UK to talk Solaris and hear the likes of Phil Harman with titles like:

64 reasons to upgrade (to Solaris 7)

Recall that 10 years ago Solaris 7 was the first release of Solaris to be 64 bit so it made the geeks laugh, well if not laugh, smile.

This month we have talks on:

OpenSolaris Security, Etude Usage, ZFS Under the Hood, SPARC new boot and much more. For those inside Sun see: https://cepedia.sfbay.sun.com/index.php?title=Uk-solaris.

Then after the (internal to Sun) meeting we can decamp to the November London OpenSolaris User Group meeting.

Wednesday Nov 14, 2007

scp, sftp, rsync to supportfiles?

As I cycled to work today I was reflecting on how easy it was to upload my webrev onto cr.opensolaris.org and contrasting that with uploading and downloading files to supportfiles.sun.com.

For those not familiar with cr.opensolaris.org when you register one opensolaris.org you have the option to upload upto 3 ssh authorised key to the site. If you are already familiar with ssh this is a simple cut'n'paste job. Eg:

ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAIEAv8aRUVQTgIqhXDJ/VAHzDEGCd3slBlAUtqjw0FytjkPeLkqPUJAQ2RBS5mN9g8IXO9uzDIZ/no0HW87J1kZhGdy/gKc/7E/z6moVG0ZWzKotfQ+AYGvH5E1WXpIpCuWPqNTEo0RMIvoGR3AwJeznKU1omQwItvQ6j+zU7cGHLZc= cg13442@estale

Then I can use scp(1), sftp(1) or rsync(1) to upload files to cr.opensolaris.org.


Now how cool would it be if that was the case for supportfiles? When our customers register they can upload the Authorised keys and then just use scp et al to manage any files they need to upload. No messing with passwords, web pages or curl. I can't help thinking system admins, who are exactly the people who upload to supportfiles, would like this.

Monday Nov 12, 2007

docs.sun.com flies

Whoooa! Whatever it was that would make access to http://docs.sun.com hang seems to have gone away.

Thank you to whoever it was.

Thursday Oct 25, 2007

So Long Matt

Yesterday the first person that I mentored at Sun left. 18 years ago Matthew Finch joined the OS group in the UK answer centre and Sun became a much better place to be. I'm sure our customers noticed it and still notice it. His dogged determination to solve the problem, find the answer, make sure the problem does not reoccur for that customer or any customer set Matthew apart. Then his knowledge of how systems work, how they are built, how they should be built if you want to diagnose them. I will miss his ability to persuade development that “yes we do need the ability JTAG scan of the system when it hangs” and I suspect so will development.

To anyone who employs Matt: Don't give Matt a problem you don't want solved.

And to Matt: Good luck in what ever you choose to do, I hope that choice includes returning to Sun.

Wednesday Oct 10, 2007

Vegas round up

This will be the last entry for CEC2007. It's been great, better than last year, I look forward to seeing the feedback from the talk Clive and I gave. It think it went well. At least I enjoyed it.

The closing day we had All hands for all the teams so I was in the one for GCS (Global Customer Services) and was even briefly on stage, sporting the yellow Tour de France Tee shirt. Then the Q&A of the executives which I think went well. I will be getting the rest of the questions that were asked and not answered due to time and posting them on cepedia somewhere so that we can get the answers.

My final thoughts on things that could make CEC2008 even better:

  • More deep technical presentations for those of us who are that way inclined.

  • We should have had a meeting for all the different TSC orgs so that we could all be in one room and put names to faces. Next year perhaps.

  • It would be good if the content builder did not give times for the presentations. Simply allows users to choose the top 20 presentations that they wish to attend numbering them 1 to 20. Then get the computer to work out the room allocations such that we get the best allocation of slots.

  • Use Sun Ray 2s with the vpn firmware installed so that they can be punched into Sun.

  • Put the Sun Rays into a FOG so that if a Sun Ray server goes down the individual Sun Rays are still usable.

  • Have some Sun Rays in Paris near the main conference room.

  • Have the party somewhere that can easily cope with the numbers rather than somewhere that can't.

They was robbed. (ZFS under the hood)

I managed to get along to hear Jason Banham and Jarod Nash doing their “ZFS under the hood bonnet presentation as the last break out before the CEC party. This was as expected a “deep dive” not for the faint hearted but they covered the material so clearly and concisely I can say this was the best break out at a CEC I have ever attended. A must for any aspiring kernel engineer.

I understand why they did not get the prize for the best presentation, some people even left before the end, but to those who already have an idea as to how a file system works and are really interested in the internals of ZFS this is a brilliant talk.

One thing to be careful of though. They gave you a chocolate if you asked a question so sit a the front, that why if anyone around you asks a question you are less likely to be hit by the chocolate as Jarod would tend to through them across the room.

Wednesday Oct 03, 2007

On my way to CEC

I'm currently in California at a meeting prior to flying out to Las Vagas for the Customer Engineering Conference 2007. I've never been to Vagas and I'm looking forward to it. I would not choose to go to Vagas but it should be interesting.

While I'm there I'll be the warm up act for Clive King who is giving a presentation called "Full Contact Debugging with the Solaris Dynamic Linker" which gives all the tricks and tips you can use to debug problems using the dynamic linker. Not really new but very useful.

If you are going to be there go to the Deep Dive sessions that are running this year. They should be very interesting two way conversations.

Thursday Sep 06, 2007

An Offer you can't refuse...

I just read this offer from Clive:

So in a one time offer only, if you are a customer who has never used SharedShell before and you want an hour or two of free remote Solaris Performance Consulting in the next week and you are happy for me to blog about it, drop me an email.

Given that Clive is one of the best performance Engineers around (he would never make such a claim himself) this is an offer that is well worth taking up. It's interesting that Clive is such a convert to the Shared Shell. The performance improvement from doing 2 hours in shared shell compared with 2 hours travelling to a site then 2 hours work and 2 hours travelling back is quite large and that assumes the site is only 2 hours away.

Thursday Aug 23, 2007

Changing names

Since name changes are the order of the day and I can't change mine to be that of one of my children (it would be confusing and how would I choose?). I've made the change I've been meaning to for some time.

My ebay id is no longer "chriscycling" but is now "triplettravel" matching my well known (cough) internet persona. I expect the bidders will now flock to my new identity.

Wednesday May 02, 2007

Open or Closed Discussions?

Here is a scenario and discussion that comes up with monotonous regularity on the technical aliases in Sun.

It starts with a question to one of the internal aliases that we have for discussions that are not appropriate for the equivalent aliases on the opensolaris.org site. These aliases were intended for “confidential” discussions, typically where sensitive customer information is in the question.

So hypothetically we have a confidential alias to discuss the fitting and maintenance of gears to bicycles confidential-bicycle-gear-interest@sun.com. There is also a bicycle-gear-interest@opensolaris.org alias. The intention is that when you have a question from a customer that is of a sensitive nature it would go to the confidential alias otherwise it goes to the external alias.

What happens next is that a question is posted to confidential-bicycle-gear-interest@sun.com that contains nothing that is obviously confidential. The response, usually from someone who does not deal directly with customers is “Why did you not post this to bicycle-gear-interest@opensolaris.org”. To which the reply comes back, “The question came from a customer and I don't want the customer to see the same question appear on opensolaris.org”. The discussion then descends into a war of attrition that is not quite as bad as “vi v emacs” or “korn shell v c shell” discussions but is more vociferous than “Shimano v Campagnolo” (mainly as cyclists will tend to stick together)1.

I can see two reasons for the “inappropriate” use of the internal aliases:

  1. The engineer does not want to expose themselves to the whole planet.

  2. The engineer is not certain that reaction of the customer would be if they saw their own question bouncing around externally, even if it by then contained no confidential information.

Point 1 is interesting as there are potential privacy issues. However since I clearly don't fall into that camp I won't comment further. If you do fall into that camp then email me to discuss or comment anonymously on this entry.

It is point 2 that I feel is not well understood. Returning to my hypothetical example:

Suppose a customer sees a commercial advantage to fitting Oval chain rings to the bikes they make so call up and ask what the implications would be for such a venture with respect to the bicycle gears. Does the engineer have the right to ask that question in the open? I say that they do not. If they did then this could potentially tip off a competitor that someone was thinking about the problem. Now should the engineer make a value call on every issue?

If the customer had been happy for the question to be asked in a public forum then they could have done this. If they wanted anonymity there are plenty of web email accounts that could be used to post. The fact that they chose to contact Sun as a commercial organisation either as part of the sales cycle or as part of a support contract means we have to respect that confidentiality 2.

On the being open or closed, I have pondered hard on whether this entry should be open or closed. If my blog goes very quiet in the next few days I guess it should have been closed!


1 The right answers are: vi, korn shell & Campagnolo;-)

2None of this is written from a legal perspective as I am not a Lawer. I don't know if we have a legal requirement to be confidential, I think we have a moral requirement which in my book makes it just as strong.

Wednesday Jan 31, 2007

Web N+1 workshop

I spent four days last week at the “Web N+1” workshop in Zurich, which has been written up by Henry Story, Peter Rieser and Dave Levy from Sun and also Bertolt Meyer who presented SkillMap and Andreas Blumauer who gave a presentation the Semantic Web from 50000m down. After these two presentations, even before we had a chance to take in the full enormity of what was presented we went on to discussing how we can move forward with our internal and external systems to give better internal tools. I'm now reflecting on what was said.

However one of the key takeaways from the workshop is that it is not documents that need to be connected as much as people. Can we use tagging in particular and applications like SkillMap to draw the real organisational structure of the organisation, as opposed to the management structure of the company. Both structures are important, however only the management structure is currently known.

Wednesday Oct 04, 2006

Executive Team 0 - Technical presentation 2

Sorry Ian, it was not much of a dilema after all.

Two cool presentations. First one on Sun's OpenSource story, then Bryan Cantrill and Brendan Gregg presenting the dtrace network provider that is being worked on. Like all things new with dtrace all you want it if not now, sooner. Very cool.

Tags:

Windows Crash Dump Analysis

For a lot of people this was the shock talk of the CEC and not because people think Windows systems don't crash but why would Sun Engineers what to know how to diagnose those. Well we get to support systems running Windows so it is useful to be able at least do a first pass analysis of the crash so that you know who to call next. Is it Windows, a driver or the hardware?

Dimitri De Wild and Feri Chua did an excellent, if a bit rushed, job of presenting the topic. This was not their fault with only forty minutes it was always going to be a rush even for just an introduction into this topic. I think I would have changed the title of the talk to “Windows Crash Dump Analysis for Solaris Kernel Engineers” though as it seemed to make some (reasonable) assumptions about the audience in some places.

Another really good talk which while I suspect I will not directly use the information it is very useful to know what Windows can do if you configure it to collect the crash dump.

Tags:

About

This is the old blog of Chris Gerhard. It has mostly moved to http://chrisgerhard.wordpress.com

Search

Archives
« April 2014
MonTueWedThuFriSatSun
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
    
       
Today