[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference 44.370::system_management

Title:system management communications forum
Moderator:CHEST::THOMPSON
Created:Fri Mar 21 1986
Last Modified:Thu Jul 08 1993
Last Successful Update:Fri Jun 06 1997
Number of topics:490
Total number of notes:2018

219.0. "Disk Fragmentation" by SHAPES::SYSTEM () Fri Mar 03 1989 09:20

	Due to severe disk fragmentation on all ADG machines.  We
	need to carry out backup/restores on each disk.  This will
	greatly improve the performance of all machines.

	To do this we will need to take the machines down for a
	period of time.  The machines would be taken down at 8pm
	each evening until the job is completed.  Below is a
	provisional schedule.  The normal backup times for each
	system remain unchanged.

	FUTURS Cluster		-	Down each evening from 8pm
					week beginning 13th March.

	CURRNT Cluster		-	Down each evening from 8pm
					week beginning 20th March.

	PASTIT Cluster		-	Down each evening from 8pm
    					week beginning 28th March.

	These backups may not take all week - it will depend on how
	many disks are on each cluster.

	Should these dates clash with urgent application work then
	please let them be known here by Wednesday 8th with dates &
	the type of work that will be taking place.

	The quicker we can repair the problem the sooner you will all
	benefit from it.

	Disk Optimization packages are currently under review.

	RA90 disks are expected to be delivered sometime in May giving
	30% increase in storage capacity.

	Regards,





    	System Management
    
T.RTitleUserPersonal
Name
DateLines
219.1HEAD::EBDONTerry Ebdon, Area Support, 781 4381Wed Mar 08 1989 10:1810
>
>	Disk Optimisation packages are currently under review.

Are you aware that VMS Engineering are strongly opposed to the use
of these packages? Apparently they have been implicated in several
disc corruptions. The ODS-2 disc structure has not been published outside
of DEC, so these products are not based on an accurate knowledge of the
disc structure and VMS disc usage.

Terry
219.2And RSTS7.2 corrupts RA81's too, said Goldilocks.NECK::THOMPSONWed Mar 08 1989 10:4918
    I have seen cheapo disk optimisation software that
    will hang a disk if it encounters an installed image file on the
    disk that it's optimising, but never one that leaves a corrupted disk
    behind. RABBIT7 is much smarter though.
    
    I've been running RABBIT7 on my GPX for a year now  and provided I
    don't attempt to touch the V5 system disk it works a treat - no sign
    of any corruption (yet). Perhaps I'm just lucky, or perhaps RD53/54 disks
    don't conform to ODS-2 standards. Seems to keep the disc I-O down a
    treat though.
    
    Are you sure this isn't another story for *real* customers like the
    'don't use any other paper except Digital's in your LN03's' one. 
    
    Have we had any disk corruption attributed to RABBIT7 on this site?
    If not, what's the cost of running RABBIT7 compared to overtime
    payments, downtime, and the tapes needed for an image backup and restore? 
    
219.3Run RABBIT, Run RABBIT, run run runSHAPES::BISHOPN1Wed Mar 08 1989 11:2211
    
    	I have heard that there can be problems when running RABBIT
    simultaneously with a VMS backup/verify. This seems to be the major
    cause of problems with this utility. With a little planning, though,
    I'm sure this situation could be avoided.
    
    	For further information on RABBIT (or DOPTER as it is known
    internal to DEC) - see the ULYSSE::DOPTER conference. Press KP7
    etc.
    
    Neil
219.4HEAD::EBDONTerry Ebdon, Area Support, 781 4381Wed Mar 08 1989 14:3215
>    Are you sure this isn't another story for *real* customers like the

YES! Please refer to the following:

	BULOVA::VAX-RMS				300.*
	VAXWRK::VMSNOTES			72.*, 152.*
	VAXWRK::NOTES$ARCHIVE::VMSNOTES_V5	2071.*
	HARDLY::SYSMGR				15.*, 151.*

These present both sides of the argument, and I would suggest you read all
of these notes before evaluating disc optimisers. In particular VAX-RMS note
300.1, entered by the VMS File Systems project manager, lists some very
important questions that should be asked. 

Terry
219.5Revised DatesSHAPES::SYSTEMThu Mar 09 1989 11:5216
    
    
    	The dates of the defragmentation on FUTURS and PASTIT have been
    	rescheduled as follows:
    
    		FUTURS		28th March
    		PASTIT		13th March
                
    
    	
    	Should thses new dates clash with urgent application work then
    please let them be known by Friday 11th March with dates & the type
    of work that will be taking place.
    
    
    Regards
219.6Revised revised datesCURRNT::SYSCONWed Mar 29 1989 20:523
    Due to the CURRNT cluster taking longer to defragment than anticipated,
    this work will continue for the week beginning 28-Mar-1989, with
    FUTURS rescheduled to 3-Apr-1989.