[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: Emergency recovery strategy needed by novice

On Fri, 7 Jan 2011, Chris Jacobs wrote:
> Dump:
> slapcat -l [ldif file]
> Add from dump, with slapd off:
> slapadd -l [ldif file]
> If you're using BDB (typical backend), you can move the contents of the dbdir specified by your config first.
> - chris


However, I can't login to the troubled server. When I try and login via
ssh, I get the password prompt, then nothing and it eventually times out
with simply, "connection lost."

I do have a connection from our application server to slapd which is
visible via 'netstat -lT' as ldaps ("ESTABLISHED") - I presume this is via
some kind of library available via Java as it's a EJB based application.

Further thoughts? Can I harness the ability to connect to get the data out
even when I don't know the first thing about the data I want to fetch,
similar to the slapcat command above?

Thanks again for any all help, comments, etc,

> Chris Jacobs, Systems Administrator
> Apollo Group  |  Apollo Marketing | Aptimus
> 2001 6th Ave Ste 3200 | Seattle, WA 98121
> phone: 206.839-8245 | cell: 206.601.3256 | Fax: 208.441.9661
> email:  chris.jacobs@apollogrp.edu
> ----- Original Message -----
> Sent: Thu Jan 06 18:32:15 2011
> Subject: Emergency recovery strategy needed by novice
> Hello All,
> I've got a bit of a problem.... Management chose a software product that
> depends on OpenLDAP's SLAPD service and had us put it into production
> without our quite understanding the software well enough to be clueful.
> The deployment is across virtual and real servers, the virtual servers
> handling the web stack and other user-oriented functionality while "real"
> servers handle data management. Regarding the data management, one system
> was chosen as a replication master, and another as slave, but being
> understaffed, we never quite got the replication of the OpenLDAP data
> configured when, today, some serious networking problems cropped up.
> The symptom is that we can establish an OpenLDAP (slapd) connection from
> the virtual side to the master, and a postgres connection - our RDBMS, but
> we can't get the rest of our primary application to connect properly so
> it's down. And it is absolutely not clear what's wrong with the master
> anyway. I can't get to the physical hardware because it's off in some data
> center somewhere that the founder located near him - and he's off
> traveling. So, we're "dead in the water."
> I was thinking that we might salvage the situation by dumping the data
> from the Postgres and OpenLDAP databases and restoring them to the slave
> system - promoting it to master - and while I've succeeded in moving over
> the Postgres data, I haven't a CLUE how to get started unloading and
> reloading the OpenLDAP data.
> ...I'm very competent with most things related to computing but I am
> _completely_ inexperienced with OpenLDAP. I've never issued any commands
> to or with OpenLDAP / slapd and have no idea what's normal for dealing
> with it. Are there dump and reload utilities?  Any and all pointers /
> help GREATLY appreciated.
> Richard