steve@violet.berkeley.edu (Steve Goldfield) (05/18/89)
Yesterday I saw a memo notifying us that Ingres would be phased out of the computer center here in Berkeley where Ingres was first developed. Apparently, the license has expired and they decided it was too much work, too expensive, and too much of a drain on hardware resources to maintain it. We are getting six-months advance warning. That doesn't mean Ingres will disappear here. Our huge database, for example, was already planning to move to a machine owned by the EECS department, which is a separate niche in the bureaucracy as far as computing equipment is concerned. However, even that is a stopgap as we expect to be out of Ingres in another year or so when a campuswide database goes into effect. I thought this development might be of more general interest. Steve Goldfield
mitchell@wdl1.UUCP (Jo Mitchell) (05/23/89)
SO - what are they replacing INGRES with? Postgres?
steve@violet.berkeley.edu (Steve Goldfield) (05/23/89)
In article <3560039@wdl1.UUCP> mitchell@wdl1.UUCP (Jo Mitchell) writes:
#> SO - what are they replacing INGRES with? Postgres?
As far as I know, it isn't being replaced by any other
database on the university's cluster of VAXes.
In fairness, I should mention that
my group was the main one using Ingres on the university
machines and probably the only one paying for it in hard
cash, so our decision to move to a cheaper department
machine was probably the principal cause of dropping
Ingres. Still, I find it ironic that here in Berkeley,
where Ingres was born, it doesn't get much respect.
From what I've gathered reading this group, there are a
couple of relevant issues. First, it appears that commercial
versions of Ingres on other machines are better-supported
(e.g., having gone to Version 6.0). Second, I assume that
in a corporate environment in which the company has outright
purchased a computer and perhaps dedicated it for Ingres' use,
exorbitant charges for cpu time are not as relevant as they
are for us in which we pay, in cash, for our usage. As long
as your equipment is powerful enough not to be brought to
its knees by Ingres and as long as Ingres does its job,
presumably you are relatively satisfied. That hasn't been
the case here.
Steve Goldfield
Industrial Liaison Program
College of Engineering
University of California at Berkeleyjas@ernie.Berkeley.EDU (Jim Shankland) (05/24/89)
In article <24756@agate.BERKELEY.EDU> steve@violet.berkeley.edu (Steve Goldfield) writes: >[My group was the principal one using, and paying for, INGRES on VAXen; >we've decided to move to a cheaper departmental computer, so the University >will no longer purchase INGRES for the VAXen.] >Still, I find it ironic that here in Berkeley, >where Ingres was born, it doesn't get much respect. > Your conclusion (INGRES doesn't get much respect) doesn't follow from the circumstances you describe. It hasn't been my experience here at Berkeley. >in a corporate environment ... exorbitant charges for cpu time are not >as relevant as they are for us in which we pay, in cash, for our usage. >As long as your equipment is powerful enough not to be brought to >its knees by Ingres and as long as Ingres does its job, >presumably you are relatively satisfied. Relational database systems have always been great resource hogs. Near as I can tell, they're all gradually getting better. More important, the days of "exorbitant charges for cpu time" are numbered (unless your need for cpu time runs into the several hundreds of MIPS). I guess we'll still be contending with each other for I/O bandwidth .... Jim Shankland jas@ernie.berkeley.edu "Blame it on the lies that killed us, blame it on the truth that ran us down"
elgie@canisius.UUCP (Bill Elgie) (05/25/89)
In article <24756@agate.BERKELEY.EDU>, steve@violet.berkeley.edu (Steve Goldfield) writes: > > #> SO - what are they replacing INGRES with? Postgres? > > As far as I know, it isn't being replaced by any other > database on the university's cluster of VAXes. So it doesn't seem like there is much need for a dbms, period. > In fairness, I should mention that my group was the main one using Ingres > on the university machines and probably the only one paying for it in hard > cash, so our decision to move to a cheaper department machine was probably > the principal cause of dropping Ingres. oh. > Still, I find it ironic that here in Berkeley, where Ingres was born, it > doesn't get much respect. Because you are not using it ? (Does this mean that if YOU decide to use something other than Berkeley UNIX, that will be "ironic" also ?) > > .... it appears that commercial > versions of Ingres on other machines are better-supported > (e.g., having gone to Version 6.0). My site and the others that I am involved with are running version 5. Sup- port seems to be fine. I haven't seen or heard anything that indicates that there is some mystical connection between support and version 6. > I assume that > in a corporate environment in which the company has outright > purchased a computer and perhaps dedicated it for Ingres' use, > exorbitant charges for cpu time are not as relevant as they > are for us in which we pay, in cash, for our usage. Sounds like you've got more than one bad guy out there to deal with. We're far from a "corporation". But that is the route we took: converted from a time sharing site's dbms to INGRES and purchased a machine for $35k to run INGRES backends. We would have done the same even if we had selected someone else's dbms (provided their distributed data base facilities worked as well): we have complex applications with complex retrievals. These can require a lot of cpu. There is no magic here, either. greg pavlov (under borrowed account), fstrf, amherst, ny
emuleomo@yes.rutgers.edu (Emuleomo) (05/26/89)
> Phasing out at Berkeley.
They probably are replacing Ingress with Progress!!
--Emuleomo O.O.