Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[TheBoy]

server, without a hot swap disk :-?

OOF's isn't

members , must put the hands in the pocket me thinks

For our needs, a 30min outage to replace a disk isn't that important.

Besides, as the server hardware comes out of my pocket, I was the one making the choice between hot plug disks, or some new tyres on the Rover

[tunnie]

No different to a regular server, at least it wasn't in the blade server we had in the lab at work. Shut the blade down, release it (just like you'd release a drive in a storage array) and slide it out.

From then on it's just a regular PC in a very small case..

Shouldn't need to shut it down. Being a blade, it'll be Enterprise class, thus 99% likely hot swap.

Hm.. our blades were obviously different, then You couldn't get at the drives without taking the casing off the blade - and since the blade is the entire server.. I can't imagine Windows would have been too happy with the power being unceremoniously removed as the blade was yanked out

Well these blades are running Red Hat, but as the whole service is fubared, the blade can be shutdown 

Built properly, a failed HDD shouldn't affect service on a server, unless they are using software mirroring....

It all went to poo when they ditched their hosting at rackspace to save money, services where brought in house and everything screwed up!

We are amazed its lasted as long as it has, I cannot get a Sybase Database to boot, it keeps corrupting chucks of the DB, the just disappear from the file system  :-?

At first it was just one, now 6 chucks have gone! Just not listed there....  :-/

We got it all back, i left it alone, came back later in day and the database had died again, booting it again chucks had disappeared!  :-/ 

[cem_devecioglu]

server, without a hot swap disk :-?

OOF's isn't

members , must put the hands in the pocket me thinks

For our needs, a 30min outage to replace a disk isn't that important.

Besides, as the server hardware comes out of my pocket, I was the one making the choice between hot plug disks, or some new tyres on the Rover

  I can guess the choice

[TheBoy]

server, without a hot swap disk :-?

OOF's isn't

members , must put the hands in the pocket me thinks

For our needs, a 30min outage to replace a disk isn't that important.

Besides, as the server hardware comes out of my pocket, I was the one making the choice between hot plug disks, or some new tyres on the Rover

  I can guess the choice

Given that Mrs TB cooks my dinner, and you lot don't, tyres for her car were always going to win

[cem_devecioglu]

No different to a regular server, at least it wasn't in the blade server we had in the lab at work. Shut the blade down, release it (just like you'd release a drive in a storage array) and slide it out.

From then on it's just a regular PC in a very small case..

Shouldn't need to shut it down. Being a blade, it'll be Enterprise class, thus 99% likely hot swap.

Hm.. our blades were obviously different, then You couldn't get at the drives without taking the casing off the blade - and since the blade is the entire server.. I can't imagine Windows would have been too happy with the power being unceremoniously removed as the blade was yanked out

Well these blades are running Red Hat, but as the whole service is fubared, the blade can be shutdown 

Built properly, a failed HDD shouldn't affect service on a server, unless they are using software mirroring....

It all went to poo when they ditched their hosting at rackspace to save money, services where brought in house and everything screwed up!

We are amazed its lasted as long as it has, I cannot get a Sybase Database to boot, it keeps corrupting chucks of the DB, the just disappear from the file system  :-?

At first it was just one, now 6 chucks have gone! Just not listed there....  :-/

We got it all back, i left it alone, came back later in day and the database had died again, booting it again chucks had disappeared!  :-/ 

 :-?

normally , the database will be marked suspect in case of some problems.. :-/

[TheBoy]

Was fsck running when it came back up?

[cem_devecioglu]

server, without a hot swap disk :-?

OOF's isn't

members , must put the hands in the pocket me thinks

For our needs, a 30min outage to replace a disk isn't that important.

Besides, as the server hardware comes out of my pocket, I was the one making the choice between hot plug disks, or some new tyres on the Rover

  I can guess the choice

Given that Mrs TB cooks my dinner, and you lot don't, tyres for her car were always going to win

thats sure.. my stomach is also the #1

[tunnie]

Was fsck running when it came back up?

Yes, after I rebooted the box i did:

Login as root
umount sda1,umount sdb1, umount sdc1

e2fsck sda1, e2fsck sdb1, e2fsck sdc1

mount –a

Login as Sybase & attempt to start, gets stuck at master device size 

[aaronjb]

No different to a regular server, at least it wasn't in the blade server we had in the lab at work. Shut the blade down, release it (just like you'd release a drive in a storage array) and slide it out.

From then on it's just a regular PC in a very small case..

Shouldn't need to shut it down. Being a blade, it'll be Enterprise class, thus 99% likely hot swap.

Hm.. our blades were obviously different, then You couldn't get at the drives without taking the casing off the blade - and since the blade is the entire server.. I can't imagine Windows would have been too happy with the power being unceremoniously removed as the blade was yanked out

What make? Never seen one without hot swap disks...

I .. can't remember I think they were HP, but I could be wrong - since I changed roles my access card doesn't even let me into the lab where the fun stuff is anymore 

[cem_devecioglu]

Was fsck running when it came back up?

Yes, after I rebooted the box i did:

Login as root
umount sda1,umount sdb1, umount sdc1

e2fsck sda1, e2fsck sdb1, e2fsck sdc1

mount –a

Login as Sybase & attempt to start, gets stuck at master device size 

did you get any error messages on fsck ?

didi you mount all volumes adressed in sybase ?

otherwise must start normally :-/

[TheBoy]

Was fsck running when it came back up?

Yes, after I rebooted the box i did:

Login as root
umount sda1,umount sdb1, umount sdc1

e2fsck sda1, e2fsck sdb1, e2fsck sdc1

mount –a

Login as Sybase & attempt to start, gets stuck at master device size 

Did you need to run fsck?  Was the database loaded before unmounting?

[TheBoy]

No different to a regular server, at least it wasn't in the blade server we had in the lab at work. Shut the blade down, release it (just like you'd release a drive in a storage array) and slide it out.

From then on it's just a regular PC in a very small case..

Shouldn't need to shut it down. Being a blade, it'll be Enterprise class, thus 99% likely hot swap.

Hm.. our blades were obviously different, then You couldn't get at the drives without taking the casing off the blade - and since the blade is the entire server.. I can't imagine Windows would have been too happy with the power being unceremoniously removed as the blade was yanked out

What make? Never seen one without hot swap disks...

I .. can't remember I think they were HP, but I could be wrong - since I changed roles my access card doesn't even let me into the lab where the fun stuff is anymore 

Can't think of an HP blade that doesn't have a SMART Array and hot plug disks.  But then I only really get involved with HP Proliants, not idea if they do Integrity based Blades...

[tunnie]

Was fsck running when it came back up?

Yes, after I rebooted the box i did:

Login as root
umount sda1,umount sdb1, umount sdc1

e2fsck sda1, e2fsck sdb1, e2fsck sdc1

mount –a

Login as Sybase & attempt to start, gets stuck at master device size 

Did you need to run fsck?  Was the database loaded before unmounting?

Those are the instructions I have, no the database was not loaded....