[lustre-discuss] Problem after implementing Lustre & Filesystem Transfer sync

[yasir at orionsolutions.co.in]

Hi,

Recently I've installed Lustre over Centos 6.7 and Lustre Client on Rocks
6.2 (Centos 6.6) after installation I'm getting following error. Need all
your help & support for below 2 Query.

 

Client Error: LustreError: 5508:0:(mdc_locks.c:918:mdc_enqueue())
ldlm_cli_enqueue: -13

OST Error: Lustre: lustre-OST0000: Connection restored to
47a64818-04a3-f5ab-ae55-5e1ff7a79eed (at 10.1.1.XX at tcp)

MDS Error: Lustre: MGS: Connection restored to
c8b97f5c-5f17-d340-de69-3d15e514ad84 (at 10.1.1.XX at tcp)

Lustre: Skipped 1 previous similar message

Lustre: MGS: Connection restored to 56b2603a-b757-f4e1-68c5-bdb91b11c634 (at
10.1.1.XX at tcp)

Lustre: Skipped 1 previous similar message

 

 

Here I would like to add one more query & need your support to clear same.

I've deploy lustre with separate server of Metadata MDT/MDS & Object OST/OSS
and mounted over lustre client machine which is having some user like: A,B,C

All infra is running on 1Gbps Ethernet Network.

 

Query: When I'm copying file near about 900Mb on lustre client in /lustre
(directory) it get copied quick but when I want to access it from any user
A.B.C it's giving permission denied error & some time directory permission
problem but after 1hrs-2hrs it allow us to access directory or files.

 

Is this just because of 1Gbps network or something else is making problem.

 

 

 

Thanks & Regards,

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20160728/0ad22668/attachment.htm>