PMA promotion records wiped by virus!

[Chad Norman]

I don't want to alarm but has anyone else heard about todays mishap at PMA?

What looked like a simple case of PMA's website being down has now turned into the possibility of PMA being wiped out by a virus!!!

Anyone heard anything?

Chad

 

[MontyPlumbs]

I don't want to alarm but has anyone else heard about todays mishap at PMA?

What looked like a simple case of PMA's website being down has now turned into the possibility of PMA being wiped out by a virus!!!

Anyone heard anything?

Chad

I find it hard to believe they don't use a backup system....seriously....even the RAF must do this!!

 

[chiprafp]

Even if true the RAF makes backups of its system at every level.

 

[mad_collie]

I don't want to alarm but has anyone else heard about todays mishap at PMA?

What looked like a simple case of PMA's website being down has now turned into the possibility of PMA being wiped out by a virus!!!

Anyone heard anything?

Chad

I heard that they've lost all the records of promotions, postings and PVRs. Just what I heard from someone 'in the know'.

The only info they have reliable access to is RAFFT results.

 

[The Masked Geek]

The system obviously crashed after failing to compute the latest illogical up banding. DT_Xtremez_14:

If PMA are on DII(F) God help us.

 

[Weebl]

I spoke to a friend of mine whose missus works at PMA and apparently they do the backup thing every 24 hours, but the backup ran automatically after the database had been corrupted. The backup just displays a string of zeros and is unusable, they 'may' have an old backup from the last day of last month but they are not quite sure.

Frantic activity apparently as old paper on everyones desks is being scrutinised to see if they can rebuild the lists.

 

[The Masked Geek]

Guess they'll start monitoring the system properly after it gets fixed then.

 

[MontyPlumbs]

I spoke to a friend of mine whose missus works at PMA and apparently they do the backup thing every 24 hours, but the backup ran automatically after the database had been corrupted. The backup just displays a string of zeros and is unusable, they 'may' have an old backup from the last day of last month but they are not quite sure.

Frantic activity apparently as old paper on everyones desks is being scrutinised to see if they can rebuild the lists.

How come they don't make a secure backup every night and keep the backups for at least a month? Even the most tin pot firm I've ever worked for did this (and the tape was placed in a fireproof safe for the month).

Mental!

 

[metimmee]

I spoke to a friend of mine whose missus works at PMA and apparently they do the backup thing every 24 hours, but the backup ran automatically after the database had been corrupted. The backup just displays a string of zeros and is unusable, they 'may' have an old backup from the last day of last month but they are not quite sure.

Frantic activity apparently as old paper on everyones desks is being scrutinised to see if they can rebuild the lists.

No way, there will be several levels of backup, they should have several days worth..if its professional IT staff managing the system....and surely for something as business critical as the HR database it will be.DT_Xtremez_09:

Maybe if they were more open about who's where, this wouldnt be such a problem. I believe REME release the full list of promotees for all to see.

 

[The Masked Geek]

Backups can only backup what is there.

A system could still run with a corrupted database and you would only know it was fecked if you tried to restart it.

The backups would have continued as normal as the backup does not care what 1s and 0s are present, it just streams them onto the tape / disk but most backup systems like Veritas will let you know when there is an issue. Unfortunately, if nobody checks the backup logs, the problem will not be spotted.

When you try to start the system up again, you find that your data is corrupt and then you need to trawl back through the tapes to find the last good backup.



I had this issue in 2003 when the snek didn't bother checking the Exchange backup logs. I got back from 2 weeks leave and had to reboot the server, which is when I noticed the problem.....We lost a couple of weeks mail that time.

 

[monobrow]

Backup strategy consisting of one rolling overnight backup? Anyone even remotely connected to looking after IT systems would know that is not bad, it's not fit for purpose. In fact, you may as well not bother.

There is either no story or something else is contributing... Besides, a virus on the RLI (regardless of location) would involve the panic button being hit hard I'd imagine....

 

[pie sandwich]

Backups can only backup what is there.

A system could still run with a corrupted database and you would only know it was fecked if you tried to restart it.

The backups would have continued as normal as the backup does not care what 1s and 0s are present, it just streams them onto the tape / disk but most backup systems like Veritas will let you know when there is an issue. Unfortunately, if nobody checks the backup logs, the problem will not be spotted.

When you try to start the system up again, you find that your data is corrupt and then you need to trawl back through the tapes to find the last good backup.

Like TMG says, Veritas (netbackup) will throw out a code should the back up fail for any reason, but it can not distinguish between a good back up and a bad one. But as a matter of course these back ups will be done on tape ( and be off sited or kept in secure storage for a month min) or if its kept on HDDs then the restore will take a few hrs and no one will notice too much fuss.

Backup strategy consisting of one rolling overnight backup? Anyone even remotely connected to looking after IT systems would know that is not bad, it's not fit for purpose. In fact, you may as well not bother.

There is either no story or something else is contributing... Besides, a virus on the RLI (regardless of location) would involve the panic button being hit hard I'd imagine....

At clueless and witless where I work, we do rolling back ups, and it works very well, but after each back up the tape media is removed and off sited to a secure compound and fresh tapes are put in for the following nights back up.

This means when we need to do a restore we just need to call back the tapes from a certain date and everything is peachy.

 

[GEM]

i went to the barbers today and he told me that the computer had caught a cold from all the whingeing buggers moaning about their HTD DT_Xtremez_14:

 

[mad_collie]

All great in theory. Have you got a replica system to run a restore at random to make sure:

1. You can actually restore to any given snapshot?

2. Your backups are actually archiving the data you need?

3. Your backup process hasn't been corrupted recently?

 

[Stuck in a rut!]

Re: PMA promotion records wiped by virus!

Ah crap, I'd heard this today in the mess and thought someone was having me on, obviously not!

I can't beleive that they would let this happen, there must be some sort of protection in place?

I wonder what the phone calls are going to be like Monday morning now, "just wondering when I'm due to be released, I was number 10", admin bod reply's "ah about that, erm, yeh, er, not anymore! Have a nice day, please call again".

I think the proverbial sh1t is about to hit the fan!

 

[monobrow]

All great in theory. Have you got a replica system to run a restore at random to make sure:

Always makes me laugh how they plan for a complete failure by moving a tape. What are you going to restore it to? A charred "now with extra carbon" server?

Backups are always a practice of assuming the data in is good. And assuming that it can extract and write over the top. It can be uber tempremental and you wouldn't know the backup is worthless until the time comes when you actually need it.

 

[mad_collie]

Always makes me laugh how they plan for a complete failure by moving a tape. What are you going to restore it to? A charred "now with extra carbon" server?

Agreed. It all depends on how the maint contract is written. In the old days, ICL would have replaced the main server kit new for old. As long as your backup for the OS & DB was watertight, you could be up & running in 48 hrs from a complete fail.

 

[RAF Bird]

Awesome I'm gonna ring them Monday morning and check that my promotion and posting to RNAS Culdrose is still on for next month.

 

[The Masked Geek]

Awesome I'm gonna ring them Monday morning and check that my promotion and posting to RNAS Culdrose is still on for next month.

I'd be asking about my promotion and posting to RAF Bahamas if I were you. =P

 

[Realist78]

Is anyone really surprised?