Bull GNU/Linux NFSv4 project

RHEL5 Robustness regression tests
March 2007

Introduction

Here are the last results for the Robustness testing between:
RHEL5(Public Gold) and linUX-2.6.19-rc6-CITI_NFS4_ALL-1 CITI patch

Tests performed

Tests have been done with the following benchs:

Test #1 : connectathon 04 (update 2003/12) (ltp-full-20060822) LTP

mode extended (-t) is used

Test #2 : nfs_fsstress (ltp-full-20060822) LTP

Tests are done with various number of subprocessess : 2, 4, 8, 16, 32

Tests are done with various length of instructions lists: 1, 2, 4, 8, 16, 32, 64, 128

Test #3 : fsx (ltp-full-20060822) LTP

fsx is run with 50000 iterations

Test #4 : Iozone (version 3.263) HERE

Run iozone -+q 30 -ace -r 64 -g 2G -i 0 -i 1 -i 2 -f /mnt/nosec/file -U /mnt/nosec

Test #5 : ffsb (version 5.2.1) FFSB

Run with default profile_stress_test file as profile

Test #6 : locks testing HERE

Run mode local done with 500 processes

Run mode network done with two clients , 500 processes each

Test #7 : dbench (3.04) HERE

Run with 10, 100, 1000 clients

Test #8 : ACL testing (ltp-full-20060822) LTP

Not Done

Robustness Results

The objective was to run:

Tests #1 #2 #3 #4 #5 #6 #7 alone for: 2 hours, 15 hours.

Tests #1 #2 #3 #4 #5 were also run all together simultaneously for: 2 hours , 15 hours.

Defaults mode is used for mount
Asynchronous mode is used in /etc/exports

The results are:

RHEL5 Robustness Results as Client in front of a 2.6.19-rc6-CITI_NFS4_ALL-1 as Server :

Client : RHEL5 (2.6.18-8.el5) on Intel(R) Xeon(TM) i386 CPU 2.80GHz bi-ways 32 bits
Server : Fedora Core6 (2.6.19-rc6-CITI_NFS4_ALL-1) on X86_64 bi-ways 64 bits

LAN Network 1Gb

Testing Tool each one alone	NFSV4 sec=sys 2 hours	NFSV4 sec=krb5 2 hours	NFSV4 sec=krb5i 2 hours	NFSV4 sec=krb5p 2 hours	NFSV4 sec=sys 15 hours	NFSV4 sec=krb5 15 hours	NFSV4 sec=krb5i 15 hours	NFSV4 sec=krb5p 15 hours
nfs_fsstress	PASSED	PASSED	PASSED	FAILED(1)	PASSED	PASSED	PASSED
FSX	PASSED	PASSED	PASSED	FAILED(1)	PASSED	PASSED	PASSED
FFSB stress file	PASSED	PASSED	PASSED	FAILED(1)	PASSED	PASSED	PASSED
IOZONE (-U)	PASSED	PASSED	PASSED	FAILED(1)
Cthon04(-t) extended	PASSED	PASSED	PASSED	FAILED(1)	PASSED	PASSED	PASSED
Locks local 500 process	PASSED	PASSED	PASSED	FAILED(1)
Locks network 1000 process 2clients	PASSED
DBENCH 10,100,1000 CLIENTS	PASSED	PASSED	PASSED	PASSED
Simultaneously fsstress +fsx +iozone (no -U) +ffsb +cthon04-t (ext)	PASSED	PASSED	PASSED		PASSED	PASSED	PASSED

Remarks:
(1) Quickly the machine totally hangs

RHEL5 Robustness Results as Server in front of a 2.6.19-rc6-CITI_NFS4_ALL-1 as Client:

server: RHEL5 (2.6.18-8.el5) on Intel(R) Xeon(TM) i386 CPU 2.80GHz bi-ways 32 bits
client : Fedora Core6 (2.6.19-rc6-CITI_NFS4_ALL-1) on X86_64 bi-ways 64 bits
LAN Network 1Gb

Testing Tool each one alone	NFSV4 sec=sys 2 hours	NFSV4 sec=krb5 2 hours	NFSV4 sec=krb5i 2 hours	NFSV4 sec=krb5p 2 hours	NFSV4 sec=sys 15 hours	NFSV4 sec=krb5 15 hours	NFSV4 sec=krb5i 15 hours
nfs_fsstress	PASSED	PASSED	FAILED (1)	FAILED (2)	PASSED	PASSED
FSX	PASSED	PASSED	FAILED (1)	PASSED(3)	PASSED	PASSED
FFSB stress file	PASSED	PASSED	FAILED (1)	PASSED	PASSED	PASSED
IOZONE (-U)	PASSED	PASSED	FAILED (1)	PASSED
Cthon04 -t extended	PASSED	PASSED	FAILED(1)	PASSED	PASSED	PASSED
Locks local 500 process	PASSED	PASSED	FAILED (1)	PASSED
Locks network 1000 process 2 clients	PASSED
dbench 10,100,1000 clients	PASSED	PASSED	FAILED (1)	PASSED
Simultaneously fsstress +fsx +iozone +ffsb +cthon04-t (ext)	PASSED	PASSED	FAILED(1)		PASSED	PASSED

Remarks:
(1) Run but without freeing the size-64 slab objects causing a memory leak. You can see it by running "slabtop -s c".
It is similar to BUGZILLA #109 which has been fixed since on the CITI patch side but not on the RHEL5 side.

(2) Connectathon Special Tests failed due to an error in 2.6.19-rc6-CITI_NFS4_ALL-1: (BUGZILLA #133)
test rewind support

write: Permission denied
special tests failed

(3) Failed with 2.6.19-rc6-CITI_NFS4_ALL-1 and passed with 2.6.21-rc4-CITI_NFS4_ALL-1 (BUGZILLA #136)

Observations 2.6.19-rc6-CITI_NFS4_ALL-1 :

We need on this side also the following patch which fixes a memory leak with krb5i ( BUGZILLA #132)

diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 2dfc4a2..b88ec10 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -740,9 +740,12 @@ static inline void nfsd4_increment_op_st

static void cstate_free(struct nfsd4_compound_state *cstate)
{
+       if (cstate == NULL)
+               return;
        fh_put(&cstate->current_fh);
        fh_put(&cstate->save_fh);
        BUG_ON(cstate->replay_owner);
+       kfree(cstate);
}

static struct nfsd4_compound_state *cstate_alloc(void)

Software versions

Linux RHEL5 Public Gold

2.6.18-8.el5

Linux Fedora Core 6	2.6.19-rc6-CITI_NFS4_ALL-1
Client userland package	util-linux-2.12 + util-linux-2.12-CITI_NFS4_ALL-3.dif
Linux nfs-utils version	nfs-utils-1.0.10 +nfs-utils-1.0.10-CITI_NFS4_ALL-3.dif
gssapi library	libgssapi-0.10
rpcsecgss library	librpcsecgss-0.14
nfsidmap library	libnfsidmap-0.18
acl library	acl_2.2.29-1 +acl-2.2.29-CITI_NFS4_ALL-3
Linux TI-RPC	0.1.7
Kerberos	V5 MIT

Hardware configuration

i386

2 processors : Intel(R)

Xeon(TM) CPU 2.80GHz, cache 512 KB

Total memory: 2Gbytes

X86_64

2 processors : Intel(R)
CPU 2.80GHz, cache 512 KB
Total memory: 2Gbytes

Network

Ethernet: 1Gb/s link

Conclusion

Between a bi-ways client and a bi-ways server, NFSV4 for RHEL5 can be considered stable for the security flavors sys, krb5 as client and server, for krb5i as client, for krb5p as server.