NT26 HardDrive issue.

[tkraffty]

yup - my client's "ker..." site just came back to life... I've told them to double-check their homepage and other regularly changing files, because I'm assuming the restore came from a backup that is atleast a day old (any details on that specifically would be appreciated!)

thomas

[gavrielh]

Is it possible that the DSN for our database connection has been blown away, our ASP sites seem to be reporting problems connecting to the database? If this is so, how do we get our DSN back for this ODBC connection?

[admin]

Hmm I will check that user now.

[admin]

http://mexicoreservations.com/ is working under that user with no issues some of them I got 500 errors. Going to restart IIS that should clear those up.

Man you got a LOT of domains under that user..

[admin]

Originally Posted by gavrielh

Is it possible that the DSN for our database connection has been blown away, our ASP sites seem to be reporting problems connecting to the database? If this is so, how do we get our DSN back for this ODBC connection?

No only the home drive had an issue not the system drive that holds your DSN's.

[generic]

come on guys, if you have mission critical ecommerce sites, that you are going to lose the customer if the site goes down for an hour or 2 or 3, etc, then why are you hosting them on a shared server setup. That is rule 1 of web hosting.

you really need to move those over to a dedicated server with raid or a backup all set to go if the drive fails, etc.

Remember this is shared hosting, there will be outages, people do run scripts that clog up the server, there are denial of service attacks, etc, etc.

In the 1.5 years I have been here I have seen and been effected by my share of them and they always get resolved, and you will not find a better hosting company that tells you more than any other host does.

Now, I am not saying that we cant expect vortech to provide a service for our dollars and I agree it should be reliable and that we should not be kept in the dark, but outages will happen and you have to be prepared for that.

Personally I would rather have the techs working on the problem than posting in the forums.

[Brangwyn]

I don't see any mention in the thread that every things completely restored etc so am I to assume that there are still people working on this server? .. I've checked a couple of domains I had on that server and they seem ok, but would like confirmation that you've finished so I can update my users, it's been almost 13 hours now since I first informed them of the issue.

[admin]

We have the backup system pushing from A-Z and I am pulling from Z-A to try and speed it up. We have a LOT of the sites working already but there is just no way to speed it up any more.

[garmol]

My domains in the server that are up so far reports incorrect Disk Usage (0 MB of 0 MB limit) in the control panel. I presumed this be fixed once the whole process of restoring the domains is completed?

[Brangwyn]

Running a pull and a push like that will most likely slow it down actually, if you think about it the drive is now having to handle multiple read/write threads and causing more seeks to be executed it's almost always faster to move data to a drive in one operation rather than trying to do multiple operations.

You may also want to consider for future restores NOT bringing the server up until it's fully completed, having some sites operational while your still restoring is also going to be dramatically slowing the entire operation down.

Please also consider just using the NT Backup Service for creating your user backups if your not already. You can write the backup file off to NAS no problem and set it to do incremental backups as well, this is actually a very effiecient way to backup and restore NT Machines that are operational.

[admin]

Almost there, 438 users restored out of 578, so its getting there. We did stop the pull of users once it finished. It did not seem to speed it up.

We tried doing backups before using the NT backup service it did not seem to work 1/2 as well plus we then have to mange 35 servers push when we can manage them all from the backup system pulling.

It does not seem to slow the server down to much with IIS running, I woul rather get sites up ASAP. I will do some testing on moving files from server to server and such to see if we can find a better way for windows restores.

[admin]

Also the CP will show zero usage as users permissions and qouta has not be added back to the new drive as of yet.

[Brangwyn]

We tried doing backups before using the NT backup service it did not seem to work 1/2 as well plus we then have to mange 35 servers push when we can manage them all from the backup system pulling.

Use NT Backup to write the file to the local drive either the user or system (assuming theres space) and then pull that file rather than pushing it to the NAS.

Have to agree with you, there has to be a better way than this, It's really a bit of a weak point in your whole operation, every time a restores been required that I can recall recently, we've basically had over 12 hours required in every case to get the server fully restored. Sure some sites have been up quicker than that, but if your the poor sucker whos account starts with a "z" thats not really any consolation.

I do think you may find the whole process runs quicker if the server is kept offline whilst it runs, and no doubt there may be improvement that can be made elswhere in the process too.

NT Backup should I think maintain the file permissions, so you wouldn't have to reset them afterwards too .. though you could always look into say scopy (NT Resource kit CD) which does move NTFS files around whilst maintain their permissions.

[jtaugher]

Originally Posted by Brangwyn

Use NT Backup to write the file to the local drive either the user or system (assuming theres space) and then pull that file rather than pushing it to the NAS.

Why not use something like Veritas Backup Exec?

In our line of business, we have many customers who rely on NT Backup and whilst we have been able to debug problems that have gone awry with NT Backup failures, we've never had a problem in a Veritas Backup Exec restore (obviously everything is monitored, and sure a backup failure isn't uncommon) -- but gosh its fast!

[jtaugher]

One of my sites that's been responding to pages for the past 25 minutes, isn't showing an database reading capability. It uses OLEDB and references e:\hshome\onestrok\ActDB\dewberryACTdb.mdb -- this directory structure hasn't changed has it? Because it's not pulling up the data results.

Also getting script time out errors. "Active Server Pages error 'ASP 0113', Script Timed Out" - "The maximum amount of time for a script to execute was exceeded."