OVH vs Hetzner dedi

[Amitz]

What to do?


For 39 euro they have an E3-1245v2 box with 16gb ECC ram and 2x 3tb enterprise drives. For my purposes the 32gb ram is nice but not really important, the ECC might make more difference (since I do long computations sometimes), and the cpu speed is about the same. I don't know if the so-called enterprise drives will be any newer or more reliable than what I have now.

I have one of those E3-1245v2 server from the auction. I posted the smartctl results here:http://lowendtalk.com/discussion/34573/smartctl-results-should-i-be-concerned


in case that you would like to see them.

 

[willie]

Thanks, I notice you didn't mention the drive model.  My drive has had a bunch of uncorrectable errors, which seems like a possible cause of concern.  My smartmon output is here:

Device Model:     ST3000DM001-9YN166                                                                  
SMART overall-health self-assessment test result: PASSED                                              
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE      
  1 Raw_Read_Error_Rate     0x000f   117   098   006    Pre-fail  Always       -       151920296      
  3 Spin_Up_Time            0x0003   093   093   000    Pre-fail  Always       -       0              
  4 Start_Stop_Count        0x0032   100   100   020    Old_age   Always       -       22             
  5 Reallocated_Sector_Ct   0x0033   100   100   036    Pre-fail  Always       -       0              
  7 Seek_Error_Rate         0x000f   086   060   030    Pre-fail  Always       -       437998817      
  9 Power_On_Hours          0x0032   086   086   000    Old_age   Always       -       12789
 10 Spin_Retry_Count        0x0013   100   100   097    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   020    Old_age   Always       -       22
183 Runtime_Bad_Block       0x0032   100   100   000    Old_age   Always       -       0
184 End-to-End_Error        0x0032   100   100   099    Old_age   Always       -       0
187 Reported_Uncorrect      0x0032   012   012   000    Old_age   Always       -       88
188 Command_Timeout         0x0032   100   099   000    Old_age   Always       -       1
189 High_Fly_Writes         0x003a   065   065   000    Old_age   Always       -       35
190 Airflow_Temperature_Cel 0x0022   065   056   045    Old_age   Always       -       35 (Min/Max 28\
/35)
191 G-Sense_Error_Rate      0x0032   100   100   000    Old_age   Always       -       0
192 Power-Off_Retract_Count 0x0032   100   100   000    Old_age   Always       -       14
193 Load_Cycle_Count        0x0032   084   084   000    Old_age   Always       -       33806
194 Temperature_Celsius     0x0022   035   044   000    Old_age   Always       -       35 (0 19 0 0)
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age   Always       -       0
240 Head_Flying_Hours       0x0000   100   253   000    Old_age   Offline      -       72657961757655
241 Total_LBAs_Written      0x0000   100   253   000    Old_age   Offline      -       5189788053836
242 Total_LBAs_Read         0x0000   100   253   000    Old_age   Offline      -       13449928820584\
0

SMART Error Log Version: 1
ATA Error Count: 52 (device log contains only the most recent five errors)
        CR = Command Register [HEX]
        FR = Features Register [HEX]
        SC = Sector Count Register [HEX]
        SN = Sector Number Register [HEX]
        CL = Cylinder Low Register [HEX]
        CH = Cylinder High Register [HEX]
        DH = Device/Head Register [HEX]
        DC = Device Command Register [HEX]
        ER = Error register [HEX]
        ST = Status register [HEX]
Powered_Up_Time is measured from power on, and printed as
DDd+hh:mm:SS.sss where DD=days, hh=hours, mm=minutes,
SS=sec, and sss=millisec. It "wraps" after 49.710 days.

Error 52 occurred at disk power-on lifetime: 8687 hours (361 days + 23 hours)
  When the command that caused the error occurred, the device was active or idle.

  After command completion occurred, registers were:
  ER ST SC SN CL CH DH
  -- -- -- -- -- -- --
  40 51 00 ff ff ff 0f  Error: UNC at LBA = 0x0fffffff = 268435455

  Commands leading to the command that caused the error were:
  CR FR SC SN CL CH DH DC   Powered_Up_Time  Command/Feature_Name
  -- -- -- -- -- -- -- --  ----------------  --------------------
  60 00 08 ff ff ff 4f 00   8d+15:00:11.505  READ FPDMA QUEUED
  ef 10 02 00 00 00 a0 00   8d+15:00:11.505  SET FEATURES [Reserved for Serial ATA]
  27 00 00 00 00 00 e0 00   8d+15:00:11.505  READ NATIVE MAX ADDRESS EXT
  ec 00 00 00 00 00 a0 00   8d+15:00:11.505  IDENTIFY DEVICE
  ef 03 46 00 00 00 a0 00   8d+15:00:11.505  SET FEATURES [Set transfer mode]


SMART Self-test log structure revision number 1
Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Extended offline    Completed without error       00%     12666         -
# 2  Extended offline    Completed without error       00%     12648         -
# 3  Extended offline    Completed without error       00%     12642         -
# 4  Extended offline    Completed without error       00%     12634         -
# 5  Extended offline    Completed without error       00%     12628         -
# 6  Extended offline    Completed without error       00%     12599         -
# 7  Extended offline    Completed without error       00%      9587         -
# 8  Extended offline    Completed without error       00%      9577         -
# 9  Extended offline    Completed without error       00%         5         -

 

[Amitz]

They are both Seagate Constellation drives (ST33000650NS).

 

[willie]

As expected Hetzner support has told me they won't replace the drive until it actually fails.  It's raid 1 but I'm still a little worried.  Also there is no monitoring of any type going to Hetzner staff (unlike OVH).  If a drive fails they won't notice unless I run my own monitoring that notifies me to notify them.   They also say they can't guarantee that a replacement drive would be in better shape.

Have you ever cancelled a Hetzner server from the robot auction?  Do you have to cancel 30 days in advance, or is immediate and prorated?  I'm thinking of getting another server and seeing if the disks are in better shape, then cancelling one or the other of them.

 

[DomainBop]

Have you ever cancelled a Hetzner server from the robot auction?  Do you have to cancel 30 days in advance, or is immediate and prorated? 

Hetzner offers a 14 day full refund period on their auction servers so you can cancel any time within the first 2 weeks and receive a full refund of your entire month's payment.

Hetzner seems to have just introduced a new server to their robot system and apparently deployed a lot of them.  i7-3770, 32gb ram, 2x 3tb drives, for 34 euro fixed price (non auction).

I picked one up.  The drives on mine are TOSHIBA DT01ACA300,  10197 power on hours on both, and no errors at all.

 

[willie]

I emailed them again pointing out an increased number of UNC errors (SMART attribute 187) and they agreed to replace the drive.  They put in another Toshiba drive similar to the one already installed, and it was new or almost-new.  After some help from #vpsboard irc guys rebuilding the RAID and an error (my fault) about Grub, I tried rebooting into the rescue system only to have that fail.  I opened a support ticket and they found the problem was a bad cpu fan, which they fixed, so my box is up and running again.  At this point I'd say Hetzner's support response has actually been pretty good.  They answered everything I sent them within an hour or so, and they turned around two hardware repairs quite quickly.

I've now transferred around a TB of data from my OVH (BHS) server to the Hetzner box, and it's kept the OVH box's 100 mbit network port saturated the whole time.  I think this reflects well mostly on OVH.  I also transferred 100GB from an Ipxcore VPS in Choopa, NJ, and got around 150mbps which is fantastic for a cheap VPS.

Overall I'm mostly happy with this box.  Certainly can't beat the price.  2x the ram, 1.5x the cpu performance, 10x the LAN bandwidth, and 1.5x the disk space, for about 2/3 the cost of my OVH server.

 

[willie]

If anyone cares, the 34 euro 32gb servers have dried up, but they seem to have quietly introduced a 16gb server (again i7-3770 with 2x 3tb drives) for 31 euro which if you don't need the 32gb ram is an even better deal, and they seem to have a steady supply of them. 

Since my 32gb box now has two almost-new HDD's I'm going to stay with it rather than downgrading to a 16gb box with drives in unknown condition, but thought I'd give a heads up.  The 16gb box is actually only 26 euro if you don't have to pay EU VAT. 

 

[GigaboxHost]

we have used Hetzner for testing and did not have much luck with performance or support.

OVH had great performance and is alway available for support.

 

[fixidixi]

@GigaboxHost

r u sure that u havnt mixed up the order?

ovh support ?

 

[willie]

I've found Hetzner's support much better than OVH's.  The hardware is also a better deal (at least for some specific configurations in the robot auction, vs. SoYouStart) and the stock availability (again for those configurations, but they are the interesting ones) is also better.  OVH seemed to have better network performance, e.g. I was able to move 2+ TB of data from OVH (BHS) to Hetzner getting the OVH box's full 100 mbit/s almost the whole time, while transfer between the same two machines in the opposite direction went at maybe 1/3 of that speed.  OVH BHS obviously also had better geographic proximity to North America but I find I'm not having any real problems using a server in Germany.

OVH seems to be repositioning itself as a higher-end provider under the OVH brand, where the budget SoYouStart line gets a few dregs as machines are freed up or decommissioned/cannibalized, but no regular new stock.  I don't know the situation with Kimsufi.  Hetzner continues to play silly pricing games with its robot auction but at least they have had plentiful supplies of this class of machines.  The one I'm currently paying 34 euro (less VAT) for now seems to go for around 41 euro most of the time, but you can sometimes see it drop a little if you're persistent.

 

[farhanideas]

I will recommend to go reviews section of Hetzner from their you will get better idea about their services and support which help you to decide

 

[splitice]

I had Hetzner servers for over 4 years, I was very impressed. Only one hard drive replacement, and it was handled professionally and without too much downtime. It would have been nice if they did hotswap replacement, but given the desktop grade hardware they sell its probably not supported.

Their support remains 100x better than OVH even on their worst days.

 

[PascM]

They are a budget provider and they perform more than good based on that.

 

[DomainBop]

Looks like Hetzner has upgraded the processors on their EX40-SSD line but hasn't updated their website yet.

Ordered an i7-4770 and 37 minutes later this was delivered:

Hardware data:

   CPU1: Intel® Xeon® CPU E3-1246 v3 @ 3.50GHz (Cores 8)
   Memory:  32059 MB
   Disk /dev/sda: 256 GB (=> 238 GiB)
   Disk /dev/sdb: 256 GB (=> 238 GiB)
   Total capacity 476 GiB with 2 Disks

Crucial CT256MX100SSD1 drives

 

[DomainBop]

Hetzner this morning...Saturday...a weekend

2:49AM ticket opened because a server is unreachable

3:25AM problem diagnosed, faulty power supply replaced, and server is back online.

36 minutes resolution for a hardware replacement... +1 

 

[TurnkeyInternet]

That is super fast for a physical repair - i'd almost think it was a power breaker/pdu tripped rather than a full on power supply  test, replace, reboot in that time- but that's sure quick!

 

[willie]

Mine had two hardware replacements (failing HDD and failed CPU fan) and it took some back and forth to convince them to replace the HDD (see my post of 04 October).  But once convinced they changed it out quickly, and they also quickly diagnosed and fixed the cpu fan.  I don't know if either was within 36 minutes but if not, both were pretty close.

 

[GreenHostBox]

We haven't hosted with Hetzner yet but we had with OVH. I wasn't satisfied with their performance and network which made us switch to a different provider.

 

[uniweb]

If you know linux, then Hetzner. super service and quality, but it was necessary for the knowledge of the administration.

 

[Joshua-Epic]

In the past we used OVH for our Europe based hosting brands but always had issues with network availability. At the present, we do have a OVH server for storage and it does the job, just not sure if you would want to use them for a public facing server.