× The internal search function is temporarily non-functional. The current search engine is no longer viable and we are researching alternatives.
As a stop gap measure, we are using Google's custom search engine service.
If you know of an easy to use, open source, search engine ... please contact support@midrange.com.


  • Subject: RE: Hey Guys - people actually think that I know what I'm talking about...
  • From: Evan Harris <spanner@xxxxxxxxxx>
  • Date: Fri, 04 May 2001 08:04:57 +1200

Brad

To be honest I meant a system rebuild :) If anyone still does these these 
days. I take it from your comments that the webserver is easily rebuilt 
somehow, and from that point of view, a recovery could be accomplished 
using scripts that rebuild things rather than restoring. Recovery and 
restore are not necessarily the same thing even though the point is often 
missed

> > 1. Have you tested a rebuild of this stuff with no problems ?
>
>Yep.  Only once.  It worked, too.  But the catch is the data on our web
>server can easily be "refreshed" from it's real sources, which are about 25
>different production machines.  So even if it was a total loss, we'd still
>be able to get it back in an hour or two.

Sound like a fair whack of your data is transient rather than "permanent".

> > I've been
> > wary of Save while active due to Al's comments.
>
>I missed these I guess.

Theres always the archives and that might be better then me trying to 
remember them. I do recall that it was to do with ensuring all applications 
were quiesced sufficiently to ensure a synchpoint was obtained across the 
saved libraries.

> > 2. How do you manage getting a synchpoint (will the web
> > server not respond
> > for a short time while this happens) ?
>
>Honestly, I don't know.  We set the backups to run at 1am (or sometime early
>am).  We don't shut down TCPIP or anything, we just run backups.  It has
>worked and saved pretty much everything.  I haven't seen any problems in the
>logs about not being able to access data, and no one has complained.  And
>trust me, they would if there was a problem.  This system used to run on NT
>and it always was bust.  Now they don't understand why it works all the
>time, and so fast.  :)

I know from experience that doing a save will not save all of the TCP stuff 
while TCP is active. I have also seen various comments that even Save while 
active will not save all the TCP stuff. Are you omitting libraries in your 
save and saving particular libraries and then rebuilding the webserver ?

This IS a valid approach IMO although as I said some people get hung up on 
save and restore instead of looking at whether its easier to rebuild rather 
than restore. If the rebuild is documented its nor different than doing a 
restore, it just uses a different command.

Personally I don't trust save regimes where I don't KNOW why something is 
working/not working. Been responsible for recovery too long I guess.

> > 3. Does using RPG/DB2 as your Backend CGI tool facilitate this in
> > comparison to all the Net.data, domino, websphere packages ?
>
>I can't say anything about domino, but with RPG I never put locks on data
>for more time than the "request" is being made.  I assume Net.Data and Java
>Servlets/JSP are the same, depending on how you write your apps.  Domino, no
>clue.  What's that?  ;)

A tool that ... ah lets not go there. I'm no fan of Domino or the people 
(in general) that seem to be involved with it :)

>Also, I am not unrealistic thinking I NEEEEEEEEED 24/7 availability or the
>world will come to an end (if anyone sites the 911 operations example, I'm
>gonna puke, in your shoes <bg>).

I have some different views on this, however, bottom line, if the customer 
demands it <shrug> ....

One point, 24/7 availability could be driven by time zone requirements - I 
just had to find a different example ;)

>I realize that the lesser of two evils (for me) is doing backups at a
>certain time, at least once a week even if it does mean folks can't view
>their order status at 2am. But then again, I'm a realist.

See comment above :)

>That and our data on our web box is simply duplicated data from other
>production machines.  So really we don't even need to do backups because of
>how our apps are done on the web machine.

Hmmmm ... don't need to do backups :) I think I covered this before. You 
obviously understand what you need to do to recover, I figure you have a 
DRP so, if getting your machine up and running is the goal and you have a 
plan, like I said, restore or rebuild - as long as it ALL comes back, whats 
the difference ?

Regards
Evan Harris

+---
| This is the WEB400 Mailing List!
| To submit a new message, send your mail to WEB400@midrange.com.
| To subscribe to this list send email to WEB400-SUB@midrange.com.
| To unsubscribe from this list send email to WEB400-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---

As an Amazon Associate we earn from qualifying purchases.

This thread ...

Replies:

Follow On AppleNews
Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].

Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.