Slow FTP in virtual Machines

[Sorin.CH]

Hi everybody

I am dealing with a VEEERRRY SLOW FTP issue on two virtual machines that are hosted on a BL860.
I have been reading all morning the threads on the matter but still could no find an answer.

The two VME are configured the same.
Uname
hpux2vm2:root >uname -a
HP-UX hpux2vm2 B.11.31 U ia64 2069479552 unlimited-user license


LANSCAN
hpux2vm2:root >lanscan
Hardware Station Crd Hdw Net-Interface NM MAC HP-DLPI DLPI
Path Address In# State NamePPA ID Type Support Mjr#
0/0/0/0 0x0EC45593FA15 0 UP lan0 snap0 1 ETHER Yes 119
0/0/2/0 0x16021F7994E8 1 UP lan1 snap1 2 ETHER Yes 119
LinkAgg0 0x000000000000 900 DOWN lan900 snap900 4 ETHER Yes 119
LinkAgg1 0x000000000000 901 DOWN lan901 snap901 5 ETHER Yes 119
LinkAgg2 0x000000000000 902 DOWN lan902 snap902 6 ETHER Yes 119
LinkAgg3 0x000000000000 903 DOWN lan903 snap903 7 ETHER Yes 119
LinkAgg4 0x000000000000 904 DOWN lan904 snap904 8 ETHER Yes 119
hpux2vm2:root >

hpux2vm2:root >lanscan -p
0
1
900
901
902
903
904


Then the lanadmin gives the following:
hpux2vm2:root >lanadmin -x 0
Speed = 1000 Full-Duplex.
Autonegotiation = On.


When I try to modify the speed it will not let me :
hpux2vm2:root >lanadmin -X 100FD 0
The option "100FD" is not valid
stats clear Clear all driver and adapter statistics
drv_pr_on Enable TCP packet reassembly in driver
drv_pr_off Disable TCP packet reassembly in driver


I have read on the forum that it might be due to a speed/duplex mismatch between the VM and the switch. How can I verify the configuration of a vswitch ?

Thanks a bunch for your answers all will be venerated.

Sorin.

[Sorin.CH]

Hi everybody, LATEST UPDATE

It seems that the "disease" might be slight different as the symptomps are different. The thing is that the FTP transfer starts okay and than after sometime it hangs...and keeps hagning.... like forever.

Then When I try to ftp again it does not work at all and only a reboot will solve the issue.

Sorin.

[Tim Nelson]

I would first look at the Virtual Host, remember all resources in an IVM or VGuest are dependant on the VHost. You cannot fix a lan issue on the VGuest if the issue is on the VHost.

-Test FTP to and from the VHost.

-Test FTPs from the VHost to the VGuest(s). There should be little issue as this never touches any wires.

Not sure why you would want to downgrade a 1000FDX to a 100FDX as you have above but again this is a virtual LAN there are no settings to set. They are all on the network backingstore supplied by the VHost.

Do some test on the VHost then report back.
Also test your source of your FTP, the problem could be there as well. Choose a couple different sources to ftp to.

[Sorin.CH]

Hi Tim,

the 1000fd-->100fd matter: I just wanted to see if I can change the speed and I have discovered that the commands does not actually work.

I have tried to ftp a file from the VM to any other host (other VMs or real outside hosts) and the same thing happens. it stops responding after a while (with different files). The strange thing is that when I try to FTP from the VM to the its real host (the bl860) the ftp finished with no problems.

Any ideas ?

Thanks

[melvyn burnard]

You cannot change the speed between a vswitch and the vNIC on a VM.
To view the vswitch, use hpvmnet -V -S vswitchname

[Sorin.CH]

Hi Melvyn,

Finally I realized that. Now still fighting with the FTP issue.

Thanks

S,

[Sorin.CH]

Hi,

Some more tests.

the environment is as follows :

HPUX-2 -real machine : 11iv2
HPUX2vm1 virtual 11iv3
HPUX2vm2 virtual 11iv3

HPUX-3 -real machine : 11iv2
HPUX3vm1 virtual 11iv3
HPUX3vm2 virtual 11iv3

The file used for testing had 2.2 GB

The tests

hpux2vm2 to hpux2vm1 OK
hpux2vm2 to hpux2 OK
hpux2vm2 to hpux3vm1 stops respondingat about 1.1GB
hpux2vm2 to hpux3vm2 stops respondingat about 1.1GB
hpux2vm2 to hpux3 stops respondingat about 1.1GB
hpux2vm2 to Windows Blades stops responding

hpux2vm1 to hpux3vm1 OK
hpux2vm1 to hpux3vm2 OK
hpux2 to hpux3 OK
hpux2 to hpux3vm1 OK

Hope this helps in finding an answer :))

Sorin.

[Steven E. Protter]

Shalom,

Check /etc/nsswitch.conf
If host resolution is files then dns

add a self resolution line to /etc/hosts and that will speed this up.

if the system name is myserver and the ip addy is 192.168.0.10

192.168.0.10 myserver myserver.yourdomain.net

SEP

[Sorin.CH]

Hi SEP,

Here is the output of my /etc/nsswitch.conf
hpux2vm2:root >cat /etc/nsswitch.conf
passwd: files
group: files
hosts: files dns
networks: files
protocols: files
rpc: files
publickey: files
netgroup: files
automount: files
aliases: files
services: files

and of /etc/hosts

127.0.0.1 localhost loopback

172.21.30.21 hpux2vm1
172.21.30.22 hpux2vm2
172.21.30.31 hpux3vm1
172.21.30.32 hpux3vm2

I still can not make it why it does not work.

S,

[Tim Nelson]

So with your tests.
hpux2vm2 to hpux2vm1 OK
hpux2vm2 to hpux2 OK
hpux2vm2 to hpux3vm1 HANGS at about 1.1GB
hpux2vm2 to hpux3vm2 HANGS at about 1.1GB
hpux2vm2 to hpux3 HANGS at about 1.1GB
hpux2vm2 to Windows Blades HANGS

hpux2vm1 to hpux3vm1 OK
hpux2vm1 to hpux3vm2 OK
hpux2 to hpux3 OK
hpux2 to hpux3vm1 OK


It looks like the main/only issue is with hpux2vm2 to other servers outside of hpux2x.

Maybe need to look at your vswitch config that hpux2vm2 is using.

From hpux2(the VHost). post output of hpvmnet -v.

Another trick with would be to run some pings off of hpux2vm2 to anywhere (not on the same Physical hosts nic) and see if you are getting some dropped packets.

or

also post lanadmin stats from hpux2vm2

[Sorin.CH]

Hi,

Well first the output of hpvmnet -V and hpvmnet -v

hpux-2:root >hpvmnet -V
Name : localnet
number : 1
PID : 1784
State : Up
Mode : Shared
NamePPA : n/a
Host MAC Address : N/A
Host IP Address : N/A
Number of guests : 0

Name : vswA
number : 4
PID : 1791
State : Up
Mode : Shared
NamePPA : lan900
Host MAC Address : 0x0017a477040c
Host IP Address : 172.21.30.20
Number of guests : 2
Guest's Name : hpux2vm1
MAC Address : 0x4201C1401D72
IP Address : 172.21.30.21
Vswitch Port : 1
Guest's Name : hpux2vm2
MAC Address : 0x0EC45593FA15
IP Address : 172.21.30.22
Vswitch Port : 2

Name : vswHBeat
number : 5
PID : 1798
State : Up
Mode : Shared
NamePPA : lan901
Host MAC Address : 0x0017a477040e
Host IP Address : 192.168.0.5
Number of guests : 2
Guest's Name : hpux2vm1
MAC Address : 0x5E845B59DF03
IP Address : 192.168.0.1
Vswitch Port : 1
Guest's Name : hpux2vm2
MAC Address : 0x16021F7994E8
IP Address : 192.168.0.3
Vswitch Port : 2

hpux-2:root >hpmvnet -v
ksh: hpmvnet: not found
hpux-2:root >hpvmnet -v
Version A.03.50.00
Name Number State Mode NamePPA MAC Address IP Address
======== ====== ======= ========= ======== ============== ===============
localnet 1 Up Shared N/A N/A
vswA 4 Up Shared lan900 0x0017a477040c 172.21.30.20
vswHBeat 5 Up Shared lan901 0x0017a477040e 192.168.0.5

Now for the lanadmin on the hpux2vm2

hpux2vm2:root >lanadmin -x 0
Speed = 1000 Full-Duplex.
Autonegotiation = On.

and the display from lanadmin/lan

Enter command: display

LAN INTERFACE STATUS DISPLAY
Thu, Sep 11,2008 09:30:35

PPA Number = 0
Description = lan0 HP PCI 1000Base-T Release B.11.31.0803
Type (value) = ethernet-csmacd(6)
MTU Size = 1500
Speed = 1000000000
Station Address = 0xec45593fa15
Administration Status (value) = up(1)
Operation Status (value) = up(1)
Last Change = 15628753
Inbound Octets = 2735819361
Inbound Unicast Packets = 5728050
Inbound Non-Unicast Packets = 1148859
Inbound Discards = 205720
Inbound Errors = 0
Inbound Unknown Protocols = 332832
Outbound Octets = 3791800333
Outbound Unicast Packets = 3858322
Outbound Non-Unicast Packets = 313
Outbound Discards = 48
Outbound Errors = 162
Outbound Queue Length = 4294967294
Specific = 655367

Press to continue


Ethernet-like Statistics Group

Index = 1
Alignment Errors = 0
FCS Errors = 0
Single Collision Frames = 0
Multiple Collision Frames = 0
Deferred Transmissions = 0
Late Collisions = 0
Excessive Collisions = 0
Internal MAC Transmit Errors = 0
Carrier Sense Errors = 0
Frames Too Long = 0
Internal MAC Receive Errors = 0


I will try now the ping part. IT is a very very strange behaviour.
Is there a trace that I might uses for the FTP transfer?
Even more I can not succedd in making the ftp log in the tranfers to /var/adm/syslog/xferlog Ven if I have the following

cat /etc/ftpd/ftpaccess
authorized users only. All activity may be monitored and reported.
log syslog+xferlog


hpux2vm2:root >cat /etc/inetd.conf | grep -i ftp
ftp stream tcp6 nowait root /usr/lbin/ftpd ftpd -l

But still after a transfer (I have tried to make one in and one out) I still do not have anything in the xferlog file.

Do I have to add also the -a -i -o ?

Thanks a bunch.

Sorin.

[Sorin.CH]

Hi All,

It seems that I finally cornered down the issue. That is
Having the environment above and running HP-UX 11.32 my BINARY MODE for FTP does not work. That is

If I try to get a file that is on the HP-UX machine(binary or asci file) using ASCII MODE it works no isses. But if the file is BINARY it will be corrupt.

Now if I try to get the same file but in BINARY MODE then the trasnfer stops responding every time at the same point for the same file.

It does not matter if I connect to the HP-UX ftp server from Windows or HP-UX. IT will stops responding anyway.

here is the status of my ftp
hpux1vm2: root> ftp 172.21.30.32
Connected to 172.21.30.32.
220 hpux3vm2 FTP server (Revision 3.0 Version wuftpd-2.6.1 Wed Dec 19 08:45:38 GMT 2007) ready.
Name (172.21.30.32:ops): sorin
230 User sorin logged in.
Remote system type is UNIX.
Using binary mode to transfer files.
ftp> status
Connected to 172.21.30.32.
No proxy connection.
Mode: stream; Type: binary; Form: non-print; Structure: file
Verbose: on; Bell: off; Prompting: on; Globbing: on
Store unique: off; Receive unique: off
Case: off; CR stripping: on
Ntrans: off
Nmap: off
Hash mark printing: off; Use of PORT cmds: on



ANY IDEAS ?

Thanks a bunch

Sorin