Is there a way to dismount a volume (with open files on it) or cause kernel panic if IO error occurs on underlying disk? OS is Linux. This is needed to prevent data corruption in case of storage availability problems.
How to back up existing software and configuration of HP Lefthand P4000 series (two P4300G2 nodes in cluster running software version 9.0) storage? We are planning to add virtual failover manager and upgrade to software version 9.5. Upgrade is planned with all iSCSI volumes disconnected from servers and we would like to be able to backup (and restore in case of upgrade failure) existing software and it's configuration (incl. replication states, etc).
I have a virtual machine running on VMware ESXi 4.1.0, machine has one Thick provisioned disk stored in VM forder (residing on local datastore). I am planning to move the virtual machine to another ESXi host with minimal downtime - I was planning to shutdown vm, make snapshot, start it up again, copy all files modified before creating snapshot to new ESXi server and some nights later when actually moving the server copy only files changed after creating snapshot.
I have created several snapshots of the machine. But always when I download vmname-flat.vmdk this file has a different MD5. I have downloaded it with vSphere Client and with Veeam fastSCP several times with each - no 2 files are the same. Why so? Is this by design?
How it is possible to copy all data till latest snapshot while machine is running and copy only latest delta while machine will be turned off?
We have multi domain Active Directory forest with a few external trusts. Let's say we have forest root domain named company.com and a few child domains in that forest - subsidiary1.com, subsidiary2.com and subsidiary3.com. We are creating firewall rules that will restrict communication to domain controllers of company.com from networks of subsidiaries.
Is there any article from Microsoft that describes required network connectivity (opened ports in firewalls) between workstations/member servers and domain controllers of other domains of the same forest required for proper operation of AD infrastructure itself?
Some information on this topic is here:
How to configure a firewall for domains and trusts
How Domain and Forest Trusts Work
However these articles don't answer my question - is access to domain controllers of forest root domain from all workstations (and member servers) of all forest domains required?
I know that practically most things (except, for example domain authentication from MacOS workstations) are working fine if DCs of forest root domain (as well as all other domains, except domain where user and computer resides) are not accessible from workstations, but I would like to look at any official information from Microsoft or to hear opinion of administrators who have long experience with running such configurations.
I am creating LDAP proxy server using OpenLDAP, server will be used for user authentication (using simple bind over SSL/TLS). Is there a way to make OpenLDAP proxy to cache simple bind credentials and respond to simple bind requests from cache in case main LDAP server goes temporary offline?
May be there are any (preferably free) alternatives that can cache authentication requests?
I have Cisco 1800 series router (running C181X-ADVIPSERVICESK9-M) with lots of IPSec tunnels. I want to add this router to OSPF and configure it to advertise routes to remote subnets accessible via IPSec tunnels (e.g. if I have tunnel VPN1 with access list that allows all traffic from my network to 10.20.30.0/24 go through that tunnel, I want router to advertise route to 10.20.30.0/24 in OSPF). Can this be achieved on Cisco IOS without explicitly adding routes to those remote subnets?