Lals,
Which node has IP address 69.10.54.190 (I assume the MDS)?
What was the iptables rule you added to open port 988 and can you adapt your
rules to log the packets you think iptables is dropping?
Also, note that the client node will connect from a "secure" port
(1023 in
this case); could another firewall rule be filtering incoming connection
attempts from such ports?
Lustre should not listen on any other ports apart from the one set by the
lnet ''accept_port'' module parameter (which defaults to 988).
Cheers,
Eric
Eric Barton Barton Software
9 York Gardens
Clifton
Bristol, BS8 4LL
United Kingdom Tel:
Mobile:
Fax:
Email: +44 (117) 330 1575
+44 (7909) 680 356
Call to arrange
eeb@bartonsoftware.com <mailto:eeb@bartonsoftware.com>
_____
From: lustre-discuss-bounces@clusterfs.com
[mailto:lustre-discuss-bounces@clusterfs.com] On Behalf Of Prajith Lal
Sent: 17 July 2006 2:01 PM
To: lustre-discuss@clusterfs.com
Subject: [Lustre-discuss] Lustre Firewall Issue!!
Hi,
I Configured Lustre FS with host1vm1 as MDS and host2vm1 as OST and host3vm1
as my client. I ran my configuration xml succefully and created the MDS,
OSTs and clients. But when I turn on my firewall it have some problems. I
updated my iptable with a new rule named lustre and open the port 988 for
lustre. This is done on both the MDS server and OST but not in the client.
This time I can successfully ran my script on these two servers but when I
try to execute the script on my client it hangs. I checked the system log
and it displays this error message
LustreError: 5256:0:(linux-tcpip.c:666:libcfs_sock_connect()) Error -113
connecting 0.0.0.0/1023 -> 69.10.54.190/988
LustreError: Connection to 69.10.54.190@tcp at host 69.10.54.190 was
unreachable: the network or that node may be down, or Lustre may be
misconfigured.
LustreError: 5256:0:(socklnd_cb.c:396:ksocknal_txlist_done()) Deleting
packet type 1 len 240 69.10.54.202@tcp->69.10.54.190@tcp
LustreError: 5256:0:(events.c:54:request_out_callback()) @@@ type 4, status
-5 req@df990e00 x18/t0 o38->mdsdev1_UUID@1vm1.test.com_UUID:12 lens 240/272
ref 2 fl Rpc:/0/0 rc 0/0
LustreError: 5278:0:(client.c:951:ptlrpc_expire_one_request()) @@@ timeout
(sent at 1153137626, 0s ago) req@df990e00 x18/t0
o38->mdsdev1_UUID@1vm1.test.com_UUID:12 lens 240/272 ref 1 fl Rpc:/0/0 rc
0/0
without firewall the script works well.
Is the Lustre FS using anyother ports for connecting with clients?
anyone have the idea? Anyhelp is appreciable..
--
Thanks
Lals
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://mail.clusterfs.com/pipermail/lustre-discuss/attachments/20060717/8dcf20a1/attachment.html