Large File Copy

Got this from a friends blog, mention of using esutil to copy large files;



http://ultankinahan.spaces.live.com/



The full article is here;



http://blogs.technet.com/askperf/archive/2007/05/08/slow-large-file-copy-issues.aspx



It mentions that eseutil is a better option for moving large files, yet someone on the above mentioned they tried to copy a 2GB file and it was slower. On the grand scheme of things I would not consider a 2GB file a large file, anything over 50GB is a large file. Migrating data from an old server to a new server generally means min file size of 10GB and normal 'Drive' backup files of 40 to 80GB at time.



I've never seen Exchange fail to move its database file from one drive to another, so I'm hopeful that tomorrow I'll try out using eseutil to move a 40GB file off a NAS onto a server and see what happens.

Swing Migration

I've been busy between different jobs recently and I keep misplacing my notes, so hopefully if I put all these things in the one blog, with some sort of standard tag I'll find them all the time.

Several issues that I had on today's swing are new, some are old, and some are just plain odd.

I had to start the w32time service
I had to start the Ismserv service

I was getting several errors in the event logs, 1722, 1815, and a few more, my notes are on site so hopefully tomorrow I'll edit this better.

The Windows Firewall started on both servers at several occasions, even though I had stopped the service.

I had to kick start several services in an attempt to get replication to start, the ones that had the biggest effect were as follows
net stop netlogon
net start netlogon
net stop ntfrs
net start ntfrs
repadmin /kcc this will get the KCC to try to start a connection and hopefully start replication

Even though I tried rebooting the swing server several times, the GC still would not start replicating for me. I went back to my notes and did the following;
Removed the GC option from the swing server, reboot the server, turn on the GC option, and then depending on if your a belt and braces person, reboot the server.
Hopefully the Directory Services event log will give you event ID 1110, which means that you may well see event 1119 5 minutes later.

Odd little bits

All stuff that you know, but all handy.

Shutting down a server or pc from another pc;
Use Computer management, either your own machines 'manage' or a mmc and computer management, choosing a remote server or pc.

Right click on the computer, advanced, startup and recovery, shutdown and REMEMBER to use restart, especially if your remote.

Sometimes, a server is locked, or otherwise madly locked up, and the above option won't work try the following;

open a command line, and type the following
shutdown /f /m \\swing (your server or pc name) /r
/f force
/m this is a remote machine, followed by \\ and then the machine name
/r shutdown and re-start.

HP G5 Servers

I've configured at least 3 HP G5 servers over the last few weeks and I've discovered that they do not like their network cards to be manually set. We generally set the NIC's to 100/ Full but when this is done, they generally refuse to talk efficiently on the LAN.

I've had this with a Allied Telison, a 3Com and a Cisco Switch, so it definitely ain't switch related.

Further on from this....this thing is breaking my heart. I rebooted the server and it was all OK, but a few hours later its back to 55 minutes to copy over 200MB of data, that's about 1Mb transfer rate. Going to check HP site and see if there is a patch or similar.