I've run a multitude of dual and single processor machines. I've never noticed a difference as far as application usability goes. Last time I ever had a problem ripping a CD and surfing the web at the same time was when I was still running my 200mmx.
Also, dual CPU will not help the "feel" of your system. If it slows down, it is because your IDE bus is busy and Windows is trying to access files off of your HD. You can increase your clock cycles and the number of processors all you want, but the PCI bus is still only 33MHz on a 32bit bus.
Dual processor makes very little difference to PC users who don't have penis size issues.
In a server, totally different story. When most OLTP databases or webservers are running, they are generally servicing more than 10 users at once.
Every switch vendor has something like this that interops fairly well. Cisco calls it Etherchannel. I think Extreme calls it channel bonding.
If you are only going to 200meg, there is no point in building new switchs and so forth when you can already go up to 400meg with channel bonding. I could be wrong there though, as I think you can bond up to 8 ports with Etherchannel.
Just hit Ebay and pick up some 2900's or 3500's and link them together.
We currently have a somewhat successful application. It runs on MS SQL and we don't like it because it takes those darn techies to make it run.
Has anyone found a silver bullet out there that will let us do really cool things for free?
We want to use dynamic content, but we don't want to do any programming, so there shouldn't be any programming involved in the product. Basically, when I leave the executives are going to be maintaining the website.
This is the way to do it. IPTables is a very powerful tool.
I'm sure if you were really interested, you could create another module that would pop up a window to prompt you if you wanted to allow the traffic through.
Just stick a rule with it at the end of your OUTPUT chain. It would catch all new application traffic and you could use that to set up new rules to be inserted above this rule.
Re:g4u source code mirror
on
Ghost for Unix
·
· Score: 2, Insightful
tar with the appropriate flags works much better. Also, if you run it on a system that doesn't write to the disk much (ie webservers) you can generally take a somewhat reliable backup without taking the system down to init 1.
Ghost is worth the money
on
Ghost for Unix
·
· Score: 5, Informative
I think it's worth it to pay for enterprise ghost and the win2k box it needs to run on if you really need ghost.
The multicast console kicks ass -- I can ghost a tonne of workstations at one time and not kill the network.
Symantecs' support infrastructure is wicked too. We haven't hit a problem that wasn't documented on their website yet.
Also, ghost understands filesystems and not raw blocks. I don't understand why reading the raw data is an advantage -- you get images the size of your hard disk or partition instead of the size of the data. Ghost 7.5 can understand fat/ntfs/ext2 and ext3. It can also do raw reads of the hard disk.
MS Project is a really complicated tool. You're probably just not seeing all the features there. I used to think the same thing, but then we started using ProjectCentral and I was also exposed to more of the UI of Project itself.
For example, certain views in Project allow you to import effort per resource per task.
Also, project central for 2000 allows you to do this, but it's a piece of crap. We're currently evaulating project server 2002. I can tell you project 2002 is a lot nicer already. Email me in a week or 2 if you want to know how 2002 fairs.
With project central I currently have all my team members update their progress every week and I use project central to dump this into the main project file. From there I can monitor what's happening, how much time is being spent where yada yada yada.
Project Central and Project Server are browser based too.
You better hope that your router has support for more VLAN's than you have users then.. if your device can only support 64 vlans, that you can only have 63 users on the network (you'll need one vlan to get to your gateway). Not to mention that you'll have to write all sorts of code to guess the subnet masq and the gateway the person is trying to use. What happens when there DNS server is on a private network and you have no idea how to route to it?
You'll have just about as much luck trying to do this as randomly crimping your cat5 cables. Any vegetable can configure DHCP -- this is an unreasonable request.
If you are interested in this sort of stuff, do a google search for surface laid fibre.
Basically, they run a line in the pavement with a pavement saw and stuff fibre down the cracks. You can do a few kilometers in a week or so -- it's fast and you don't have to trench.
and so forth. You can then add files as you want -- if you don't want to back up everything in/etc as it was created automagically, then don't.
This set up uses a very mature piece of software. It has a lot of nice interfaces (cvsweb, wincvs). It also gives your users the ability to pull their own backups and doing branching of their home directories:)
It works really well as you can back up the central repository at what ever frequency you want.
It may take a lot of storage to back it all up, but it will probably be smaller than an oracle database and you will get diffs for your history.
I've never seen a CVS repository get corrupted -- I've seen working copies get destroyed or problems with permissions.
The important thing to remember is that you can always fix CVS -- if you do get a corruption it most likely will only be in one file. You could probably open it with vi and fix it too (as long as it isn't zero length, which would happen if your drive ran out of space).
Of course, even a large CVS repository tars down pretty tightly, so you can keep lots of backups too.
Take a look at NetFilter/Iptables. It is the new firewall code for linux 2.4.
It is supposed to be an extensible framework that modules for various types of filtering can be written for. For example, you can filter outgoing messages by uid.
Using this, it shouldn't be too difficult to write a library to do this.
First off, *do not use mailman*. This is easily the worst mailing list software you can find. It mails passwords, clear text, monthly. I think gnu mailmain is even two steps worse than majordomo, which is hard to complish.
Qmail is fast, efficient, easy to configure (the config files make sense) and there is a huge amount of support.
Qmail has been tight for years; the code hasn't changed in a long time. The only problem with that is the documentation is out of date. I heard rumours of a qmail 1.04 to fix the documentation.
After you choose qmail, I recommend ezmlm for mailing lists.
The configuration that always comes through for me is:
qmail+vpopmail+ezmlm
Make sure all of your domains are vpopmail virtuals.
Also, if you are affraid to get into the hardcore configuration right away, start with qmailadmin. It supports all the ezmlm stuff, so you can use the gui right away. You'll start running into limitations, but you will have lots of examples to use from that point.
btw, I have unsubscrbed from mailing lists because they use mailman.
Everyone else has been yelling about WINS, so I won't say anything about that.
I recommend getting an L3 Switch (which is essentially a router with only RJ45 Jacks).
Depending on how many subnets you want, Cisco has one for $10,000 which has 48ports, or SMC has a 16 port version for around $2000. I put the SMC Tigerswitch on our network and it works great.
An L3 switch works by creating vlans that can be subnets. The main difference between this and an L2 switch is that it bridges the vlans when traffic is destined for another subnet (routing).
Actually, they do this to make you buy more. The funny thing is that people don't buy based on the best price, they buy on the catchiest price. Also, if something is more expensive it is 'better'.
This doesn't work for everything, but that is why they test prices -- to find the price that they will sell the most of a particular product at. Sometimes pricing stuff higher works better. They are not doing this to rip you off (ie raise the prices when you want it based on your purchase history).
In reality, when a company like amazon sees that you have been purchasing a lot of stuff, they want you to purchase more so they are going to send you coupons and special offers and such. They want you to buy and 'ripping you off' is not going to make you buy..
Ummm, you don't need to do the dns thing; you can also use mod_rewrite on a cluster of faster computers to do the redirecting to the back end..
And if you don't use virtual hosts on the back end, you could have the rewrite box forward the request to serverid.domain.com which always rewrites to www.domain.com, you can seamlessly reallocate locations.
The thing most people forget is that banner ads aren't to help support websites who need funding, but to generate traffic to the website that is paying money for the ads. Therefore, paying to not see them doesn't make sense.
To most companies, a single sale through a banner ad isn't worth the effort in the long run; it is generating a back end to push other products to. Therefore, a micropayment to not see ads is worth nothing to the top line growth of a company as the consumer base does not grow.
So whos fault is it for leaving your window unlocked? I'm not saying it is right to hop in and take all your goods or destroy anything, but that isn't how other people think.
When one of my production servers get scanned, I reverse scan them for a fingerprint and to try and determine if they are a box that got hacked ( usually solaris with rpc running:) ) or just someone who happened to port scan me.
I recommend qmail in its place. Using it, you can put all of your dialup user's ips. This is assuming that you are the one handing out IP's -- you will have a specific block of them, so you can force that you only relay from those hosts.
Also, don't use sendmail. It stores all of your emails in one big file. What happens when you get a mailbox file that is 70-700megs big? When pop comes along, it starts timing itself out when you copy the box from username to.username.pop and you'll kill your pop server.
qmail stores each email in a seperate file to prevent this. If you have all the wrapper programs it runs under give the process the resources it needs, you can easily store gigabytes 'in your pop account'.
Not to mention all the benefits of vpopmail.
I could probably go on and on about why you shouldn't use sendmail.. so I'll stop now.
This has to do with the way it updates its tables. Do not connect access to a live server. I've seen this with MySQL, but I imagine that it is exactly the same with every other dbms.
Consider yourself warned.
MySQL vs MySQL & mSQL (oreilly)
on
MySQL
·
· Score: 2
As an owner of both books, I'd have to say that the MySQL book is far better than the OReilly version.
I've only had a chance to flip through the MySQL book, but just in that, far more has been covered than in the OReilly one. Not only that, but it seems to be missing the technical and grammatical flaws of the My&m.
This is a strange feeling, but for once I will have to recommend the non-oreilly version of a book.
Slashdot Mirror
I've run a multitude of dual and single processor machines. I've never noticed a difference as far as application usability goes. Last time I ever had a problem ripping a CD and surfing the web at the same time was when I was still running my 200mmx.
Also, dual CPU will not help the "feel" of your system. If it slows down, it is because your IDE bus is busy and Windows is trying to access files off of your HD. You can increase your clock cycles and the number of processors all you want, but the PCI bus is still only 33MHz on a 32bit bus.
Dual processor makes very little difference to PC users who don't have penis size issues.
In a server, totally different story. When most OLTP databases or webservers are running, they are generally servicing more than 10 users at once.
Not really.
Consider that the ram drive, if using an ATA or SCSI interface would still plug into the PCI bus as some point.
Every switch vendor has something like this that interops fairly well. Cisco calls it Etherchannel. I think Extreme calls it channel bonding.
If you are only going to 200meg, there is no point in building new switchs and so forth when you can already go up to 400meg with channel bonding. I could be wrong there though, as I think you can bond up to 8 ports with Etherchannel.
Just hit Ebay and pick up some 2900's or 3500's and link them together.
Has anyone found a silver bullet out there that will let us do really cool things for free?
We want to use dynamic content, but we don't want to do any programming, so there shouldn't be any programming involved in the product. Basically, when I leave the executives are going to be maintaining the website.
Can anyone help me?
I'm sure if you were really interested, you could create another module that would pop up a window to prompt you if you wanted to allow the traffic through.
Just stick a rule with it at the end of your OUTPUT chain. It would catch all new application traffic and you could use that to set up new rules to be inserted above this rule.
tar with the appropriate flags works much better. Also, if you run it on a system that doesn't write to the disk much (ie webservers) you can generally take a somewhat reliable backup without taking the system down to init 1.
The multicast console kicks ass -- I can ghost a tonne of workstations at one time and not kill the network.
Symantecs' support infrastructure is wicked too. We haven't hit a problem that wasn't documented on their website yet.
Also, ghost understands filesystems and not raw blocks. I don't understand why reading the raw data is an advantage -- you get images the size of your hard disk or partition instead of the size of the data. Ghost 7.5 can understand fat/ntfs/ext2 and ext3. It can also do raw reads of the hard disk.
btw, I don't work for symantec.
For example, certain views in Project allow you to import effort per resource per task.
Also, project central for 2000 allows you to do this, but it's a piece of crap. We're currently evaulating project server 2002. I can tell you project 2002 is a lot nicer already. Email me in a week or 2 if you want to know how 2002 fairs.
With project central I currently have all my team members update their progress every week and I use project central to dump this into the main project file. From there I can monitor what's happening, how much time is being spent where yada yada yada.
Project Central and Project Server are browser based too.
You're probably looking for pronto.. and it can use a few different database back ends. It works well, just don't use the CSV driver :)
You'll have just about as much luck trying to do this as randomly crimping your cat5 cables. Any vegetable can configure DHCP -- this is an unreasonable request.
Basically, they run a line in the pavement with a pavement saw and stuff fibre down the cracks. You can do a few kilometers in a week or so -- it's fast and you don't have to trench.
We don't use Filers where I work, but just the local file system. Our CVS root looks something like this:
systems/$hostname
systems/$hostname/home
systems/$hostname/etc
systems/$hostname/var
and so forth. You can then add files as you want -- if you don't want to back up everything in /etc as it was created automagically, then don't.
This set up uses a very mature piece of software. It has a lot of nice interfaces (cvsweb, wincvs). It also gives your users the ability to pull their own backups and doing branching of their home directories :)
It works really well as you can back up the central repository at what ever frequency you want.
It may take a lot of storage to back it all up, but it will probably be smaller than an oracle database and you will get diffs for your history.
The important thing to remember is that you can always fix CVS -- if you do get a corruption it most likely will only be in one file. You could probably open it with vi and fix it too (as long as it isn't zero length, which would happen if your drive ran out of space).
Of course, even a large CVS repository tars down pretty tightly, so you can keep lots of backups too.
I think the world needs a place where are the console jockies can hang out and wait for 'biz'.
Hey,
Take a look at NetFilter/Iptables. It is the new firewall code for linux 2.4.
It is supposed to be an extensible framework that modules for various types of filtering can be written for. For example, you can filter outgoing messages by uid.
Using this, it shouldn't be too difficult to write a library to do this.
First off, *do not use mailman*. This is easily the worst mailing list software you can find. It mails passwords, clear text, monthly. I think gnu mailmain is even two steps worse than majordomo, which is hard to complish.
Qmail is fast, efficient, easy to configure (the config files make sense) and there is a huge amount of support.
Qmail has been tight for years; the code hasn't changed in a long time. The only problem with that is the documentation is out of date. I heard rumours of a qmail 1.04 to fix the documentation.
After you choose qmail, I recommend ezmlm for mailing lists.
The configuration that always comes through for me is:
qmail+vpopmail+ezmlm
Make sure all of your domains are vpopmail virtuals.
Also, if you are affraid to get into the hardcore configuration right away, start with qmailadmin. It supports all the ezmlm stuff, so you can use the gui right away. You'll start running into limitations, but you will have lots of examples to use from that point.
btw, I have unsubscrbed from mailing lists because they use mailman.
Hey,
Everyone else has been yelling about WINS, so I won't say anything about that.
I recommend getting an L3 Switch (which is essentially a router with only RJ45 Jacks).
Depending on how many subnets you want, Cisco has one for $10,000 which has 48ports, or SMC has a 16 port version for around $2000. I put the SMC Tigerswitch on our network and it works great.
An L3 switch works by creating vlans that can be subnets. The main difference between this and an L2 switch is that it bridges the vlans when traffic is destined for another subnet (routing).
have fun.
Actually, they do this to make you buy more. The funny thing is that people don't buy based on the best price, they buy on the catchiest price. Also, if something is more expensive it is 'better'.
This doesn't work for everything, but that is why they test prices -- to find the price that they will sell the most of a particular product at. Sometimes pricing stuff higher works better. They are not doing this to rip you off (ie raise the prices when you want it based on your purchase history).
In reality, when a company like amazon sees that you have been purchasing a lot of stuff, they want you to purchase more so they are going to send you coupons and special offers and such. They want you to buy and 'ripping you off' is not going to make you buy..
Ummm, you don't need to do the dns thing; you can also use mod_rewrite on a cluster of faster computers to do the redirecting to the back end..
And if you don't use virtual hosts on the back end, you could have the rewrite box forward the request to serverid.domain.com which always rewrites to www.domain.com, you can seamlessly reallocate locations.
The thing most people forget is that banner ads aren't to help support websites who need funding, but to generate traffic to the website that is paying money for the ads. Therefore, paying to not see them doesn't make sense.
To most companies, a single sale through a banner ad isn't worth the effort in the long run; it is generating a back end to push other products to. Therefore, a micropayment to not see ads is worth nothing to the top line growth of a company as the consumer base does not grow.
So whos fault is it for leaving your window unlocked? I'm not saying it is right to hop in and take all your goods or destroy anything, but that isn't how other people think.
:) ) or just someone who happened to port scan me.
When one of my production servers get scanned, I reverse scan them for a fingerprint and to try and determine if they are a box that got hacked ( usually solaris with rpc running
Is this from experience, or because somebody told you that beos is good for multimedia?
Sendmail is ancient. Stop using it.
.username.pop and you'll kill your pop server.
I recommend qmail in its place. Using it, you can put all of your dialup user's ips. This is assuming that you are the one handing out IP's -- you will have a specific block of them, so you can force that you only relay from those hosts.
Also, don't use sendmail. It stores all of your emails in one big file. What happens when you get a mailbox file that is 70-700megs big? When pop comes along, it starts timing itself out when you copy the box from username to
qmail stores each email in a seperate file to prevent this. If you have all the wrapper programs it runs under give the process the resources it needs, you can easily store gigabytes 'in your pop account'.
Not to mention all the benefits of vpopmail.
I could probably go on and on about why you shouldn't use sendmail.. so I'll stop now.
Warning:
ODBC puts a massive strain on a database server!
This has to do with the way it updates its tables. Do not connect access to a live server. I've seen this with MySQL, but I imagine that it is exactly the same with every other dbms.
Consider yourself warned.
As an owner of both books, I'd have to say that the MySQL book is far better than the OReilly version.
I've only had a chance to flip through the MySQL book, but just in that, far more has been covered than in the OReilly one. Not only that, but it seems to be missing the technical and grammatical flaws of the My&m.
This is a strange feeling, but for once I will have to recommend the non-oreilly version of a book.