Page 6 of 11 FirstFirst ... 3456789 ... LastLast
Results 126 to 150 of 270
  1. #126
    Join Date
    May 2012
    Location
    Germany
    Posts
    264
    Quote Originally Posted by ThePriest View Post
    The second server was unstable too, but no messages in kernel log. This one I've decided to cancel and keep my old ex4s for awhile.

    Now I'm going to wait for ex6-ssd.
    Hetzner have great configurations for awesome price, I hope they release enterprise version quickly.
    I don't even need that new CPUs, would be satisfied with current ex6s + ssd + new bandwidth.
    I'm sorry to hear you're also having problems, though it sounds like in the case of your first server it was a cooling issue that has since been fixed. I hope that server doesn't crash again. If it does you can also ask our technicians about a BIOS update.

    As for the "new" EX6S, we are still planning that and it's good to know you'd be interested in an SSD version. Thanks for letting us know!

  2. #127
    Quote Originally Posted by Hetzner_Rep View Post
    Hello laurentd, I'm sorry to hear about the issues you are having with our new EX40 server. Thank you for posting in this thread and thank you for the tone of your post, it's quite refreshing.

    We have gotten feedback from a small group of clients with similar problems as yours. Our preliminary analysis has indicated that there seems to be a compatibility issue with a certain shipment of memory modules.

    The RAM modules we use are officially listed by the Manufacturer as being compatible with the Motherboard. This is also something we tested rigorously before offering the servers, and we found no problems. However, since then we have gotten a batch of (the same) modules that use a different memory die, and this is apparently what is causing problems. We are in in contact with both MSI and Kingston to identify these modules and to resolve this issue.

    MSI has been able to provide us with a BIOS update that should resolve the problem. You mentioned a BIOS update was already done, but it could be that that update was of a previous version. Please get in touch with our technicians and ask them to check the BIOS again, to make sure you have received the most up to date version.

    Thanks for your quick answer. I have just made a HW diff on all our EX40, and it seems in fact that the server with the beta BIOS did not crash since
    the upgrade (I assumed so because we restarted it several times since the upgrade and I assumed it had crashed again.)

    The beta BIOS is V2.4B3, released on 07/10/2013 . I'll make sure this is the newest one and will request an upgrade on our other servers.

    I will post an update on this topic once we have enough data to make a new feedback.

  3. #128
    Join Date
    May 2012
    Location
    Germany
    Posts
    264
    Quote Originally Posted by laurentd View Post
    Thanks for your quick answer. I have just made a HW diff on all our EX40, and it seems in fact that the server with the beta BIOS did not crash since
    the upgrade (I assumed so because we restarted it several times since the upgrade and I assumed it had crashed again.)
    That's great to hear! As mentioned before, a BIOS update seems to solve the issue, so I'm glad that is the case with you.

    The beta BIOS is V2.4B3, released on 07/10/2013 . I'll make sure this is the newest one and will request an upgrade on our other servers.
    Yeah, that is the BIOS update we're currently giving out that should fix the issue.

    I will post an update on this topic once we have enough data to make a new feedback.
    Let's hope it's a positive update

  4. #129
    Join Date
    Feb 2006
    Location
    ::1/128
    Posts
    247
    Quote Originally Posted by Hetzner_Rep View Post
    That's great to hear! As mentioned before, a BIOS update seems to solve the issue, so I'm glad that is the case with you.



    Yeah, that is the BIOS update we're currently giving out that should fix the issue.



    Let's hope it's a positive update

    I hope for a positive outcome too.

    ex40-ssd crashed on the first 6 hours. Only centos 64bit installed (minimal). Didn't even had time for any other configurations (apache / bind / panel / and so on).

    I've sent a ticket, replied with
    "as requested we've checked the and optimized the BIOS-settings".
    that was 4 hours ago. In these 4 hours crashed 3 times again.

    Second time I got a reply
    "Unfortunately the server didn't showed a screen output and didn't respond to
    keystrokes. Because of this we can't tell you the reason of the issue."
    Sent 3rd ticket response about crashing again without touching anything.

    Hope for a fix. If something can't even has screen output or any keystroke doesn't work, I hardly believe it's a software issue.

  5. #130
    Join Date
    May 2009
    Location
    Markham, Canada
    Posts
    458
    In our case BIOS V2.14B3 *seems* to have resolved this. 20+ hours of uptime so far. Can only tell for sure after 2-3 days have passed.

    We use basic centos-6 64-bit installation too.
    Dan
    Check out our premium self-managed SSD VPS

  6. #131
    Join Date
    Jun 2009
    Posts
    1,219
    I also see those messages in kernel.log:

    Aug 3 12:35:07 node kernel: CPU1: Core power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU4: Core power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU7: Core power limit notification (total events = 2)
    Aug 3 12:35:07 node kernel: CPU0: Core power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU5: Core power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU3: Core power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU6: Core power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU2: Core power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU4: Package power limit notification (total events = 3)
    Aug 3 12:35:07 node kernel: CPU7: Package power limit notification (total events = 1)
    Aug 3 12:35:07 node kernel: CPU0: Package power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU5: Package power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU3: Package power limit notification (total events = 4)
    Aug 3 12:35:07 node kernel: CPU6: Core power limit normal
    Aug 3 12:35:07 node kernel: CPU2: Core power limit normal
    Aug 3 12:35:07 node kernel: CPU4: Core power limit normal
    Aug 3 12:35:07 node kernel: CPU7: Core power limit normal
    Aug 3 12:35:07 node kernel: CPU0: Core power limit normal
    Aug 3 12:35:07 node kernel: CPU5: Core power limit normal
    Aug 3 12:35:07 node kernel: CPU3: Core power limit normal
    Aug 3 12:35:07 node kernel: CPU4: Package power limit normal
    Aug 3 12:35:07 node kernel: CPU7: Package power limit normal
    Aug 3 12:35:07 node kernel: CPU0: Package power limit normal
    Aug 3 12:35:07 node kernel: CPU5: Package power limit normal
    Aug 3 12:35:07 node kernel: CPU3: Package power limit normal
    Aug 3 12:35:07 node kernel: CPU1: Core power limit normal
    Do I have to worry? It is a new EX40-SSD...

  7. #132
    Join Date
    May 2009
    Location
    Markham, Canada
    Posts
    458
    Quote Originally Posted by Amitz View Post
    I also see those messages in kernel.log:



    Do I have to worry? It is a new EX40-SSD...
    Sort of... What kind of uptimes are you getting?
    Dan
    Check out our premium self-managed SSD VPS

  8. #133
    Join Date
    Jun 2009
    Posts
    1,219
    Well, I just had the server for some hours until it suddenly went down. They did a 12 hours hardware check and found no issue. They brought the server back and it is running well since then. But we are only talking about approx. 3 days without problems.

    root@node ~ # uptime
    19:51:39 up 3 days, 16:50, 1 user, load average: 0.34, 0.28, 0.27
    root@node ~ #
    So should those messages worry me? I never came across them and I am obviously not tech-savy enough to judge...

  9. #134
    Join Date
    Jun 2009
    Posts
    1,219
    Okay, Update: I opened a ticket with Hetzner about the above kernel.log events and received a very friendly answer within 30mins by a Support Tech. I was told that this error message results from a kernel bug and that it could be fixed by adding
    -----------------%<-----------------
    clearcpuid=229
    -----------------%<-----------------
    as Kernel parameter within Grub.

    I then asked whether this change would be crucial and was told (< 5mins) that this messages are non-critical, followed by a brief explanation in german that I am too lazy to translate, but that have been helpful. I am quite impressed by the speed and friendliness of their support on a weekend...

  10. #135
    Join Date
    May 2009
    Location
    Markham, Canada
    Posts
    458
    clearcpuid=229
    What this does it it disables "pln" (power limit notification) feature of the CPU. So basically you are stopping those notifications, however, this does not mean the underlying issue is solved (i.e. reason for those notifications).

    When our EX40 server was upgraded to BIOS V2.14B3, these messages stopped WITHOUT disabling PLN, which is a very promising result. Server did not crash either, but we are waiting for 3-5 days before we can call this a solution.
    Dan
    Check out our premium self-managed SSD VPS

  11. #136
    Join Date
    Jun 2009
    Posts
    1,219
    Well, damn...
    My EX40-SSD became unresponsive again for 1.5 and another 1 hour this night. So 2.5 hours of downtime.
    I have no idea why. I will still try to work with their support because they were quick & friendly until now, but I am about to loose faith in that new server...

  12. #137
    Join Date
    May 2009
    Location
    Markham, Canada
    Posts
    458
    Ask them to update BIOS in your server to V2.14B3, looks like it fixes these issues.
    Dan
    Check out our premium self-managed SSD VPS

  13. #138
    Quote Originally Posted by dcdan View Post
    Ask them to update BIOS in your server to V2.14B3, looks like it fixes these issues.
    Hi,

    V2.14B3 or V2.4B3 ?!

    We have V2.14B3 and still experiencing the same issue with 2 of our servers.
    KavHost - Expand the web
    NOT Oversold Shared and Reseller hosting
    PREMIUM VPS and Dedicated servers

  14. #139
    Join Date
    May 2009
    Location
    Markham, Canada
    Posts
    458
    This version number is from dmidecode output: V2.14B3

    So far we have 47 hours of uptime (before that we were getting 8-14 before the crash).
    Dan
    Check out our premium self-managed SSD VPS

  15. #140
    Join Date
    Feb 2006
    Location
    ::1/128
    Posts
    247
    Quote Originally Posted by dcdan View Post
    Ask them to update BIOS in your server to V2.14B3, looks like it fixes these issues.
    They were kind enough to update bios.
    After 8 hours of uptime it crashed again.
    Stock centos 6.4 without any load (0 users in it).

  16. #141
    Join Date
    Jun 2009
    Posts
    1,219
    They "prophylactically" swapped the RAM in my EX40-SSD now. While I tend to cancel it already because I do not have the time and will to chase errors with the server that I have not caused, I will give it a chance for some days. Let's see if it crashes again. That is really unfortunate - I like Hetzner's panel and their support was quick and friendly too throughout the weekend...

  17. #142
    Join Date
    Nov 2009
    Location
    /etc/my.cnf
    Posts
    10,657
    Quote Originally Posted by Amitz View Post
    They "prophylactically" swapped the RAM in my EX40-SSD now. While I tend to cancel it already because I do not have the time and will to chase errors with the server that I have not caused, I will give it a chance for some days. Let's see if it crashes again. That is really unfortunate - I like Hetzner's panel and their support was quick and friendly too throughout the weekend...
    They've already said it was due to a ram issue with them getting a different chip on the modules which is entirely different to the ones they tested first of all, This happens quite regularly hence you need to be careful when purchasing ram in large quantities.
    UK Based Proactive Server Management.
    Zabbix Enterprise 24/7 Monitoring.

  18. #143
    Join Date
    May 2009
    Location
    Markham, Canada
    Posts
    458
    We have the following RAM modules in our server (from dmidecode):
    Kingston 99U5471-037.A00LF
    Configured Clock Speed: 1333 MHz
    Dan
    Check out our premium self-managed SSD VPS

  19. #144
    Join Date
    Jun 2009
    Posts
    1,219
    Here, after the Swap:

    Kingston 99U5471-037.A00LF
    Configured Clock Speed: 1600 MHz

  20. #145
    Quote Originally Posted by dcdan View Post
    We have the following RAM modules in our server (from dmidecode):
    Kingston 99U5471-037.A00LF
    Configured Clock Speed: 1333 MHz
    Quote Originally Posted by Amitz View Post
    Here, after the Swap:

    Kingston 99U5471-037.A00LF
    Configured Clock Speed: 1600 MHz
    The same modules different Clock Speed ?!
    KavHost - Expand the web
    NOT Oversold Shared and Reseller hosting
    PREMIUM VPS and Dedicated servers

  21. #146
    Join Date
    Jul 2010
    Posts
    38
    Quote Originally Posted by Amitz View Post
    Here, after the Swap:

    Kingston 99U5471-037.A00LF
    Configured Clock Speed: 1600 MHz
    The same as i have had, all the time. My bios was upgraded yesterday, 21 hours uptime now.

  22. #147
    OK here is what i have notice.

    We have 2 servers SSD version.

    Both servers have the V2.14B3 version on it and both server ware heaving problems.

    I open ticket for one of the servers to perform the bios check/upgrade/configuration and after they return the server i have the same BIOS version but i notice that the RAM clock speed is downgraded from 1600MHz to 1333MHz.
    KavHost - Expand the web
    NOT Oversold Shared and Reseller hosting
    PREMIUM VPS and Dedicated servers

  23. #148
    Join Date
    May 2012
    Location
    Germany
    Posts
    264
    Hello everyone, here's a general response to the questions and comments posted in this thread.

    If the BIOS update doesn't fix the issue, then the "issue" is most likely not with the RAM, but is either some other hardware problem or a software problem. Please try to make sure it is not a software problem yourself. If the server freezes or crashes while it is in the Rescue System then it is most likely a hardware problem. Our technicians can do a hardware check and will obviously replace any faulty hardware.

    One of the fixes we found that helped (before we got the BIOS update) was changing the speed of the RAM to 1333Mhz. If you have gotten a BIOS update the RAM can be run at 1600Mhz.
    Also, all new servers we provide already have the latest BIOS, so they shouldn't have any problems.

    The "power limit" notifications are basically a bug, and can be ignored or removed. See also the following bug report (comment 35 is helpful):
    https://bugzilla.kernel.org/show_bug.cgi?id=36182

  24. #149
    Quote Originally Posted by Hetzner_Rep View Post
    Hello everyone, here's a general response to the questions and comments posted in this thread.

    If the BIOS update doesn't fix the issue, then the "issue" is most likely not with the RAM, but is either some other hardware problem or a software problem. Please try to make sure it is not a software problem yourself. If the server freezes or crashes while it is in the Rescue System then it is most likely a hardware problem. Our technicians can do a hardware check and will obviously replace any faulty hardware.

    One of the fixes we found that helped (before we got the BIOS update) was changing the speed of the RAM to 1333Mhz. If you have gotten a BIOS update the RAM can be run at 1600Mhz.
    Also, all new servers we provide already have the latest BIOS, so they shouldn't have any problems.

    The "power limit" notifications are basically a bug, and can be ignored or removed. See also the following bug report (comment 35 is helpful):
    https://bugzilla.kernel.org/show_bug.cgi?id=36182
    Hi,

    What is the correct problem-free BIOS?
    V2.14B3 or V2.4B3 ?!
    KavHost - Expand the web
    NOT Oversold Shared and Reseller hosting
    PREMIUM VPS and Dedicated servers

  25. #150
    Join Date
    May 2009
    Location
    Markham, Canada
    Posts
    458
    Quote Originally Posted by Hetzner_Rep View Post
    The "power limit" notifications are basically a bug, and can be ignored or removed. See also the following bug report (comment 35 is helpful):
    https://bugzilla.kernel.org/show_bug.cgi?id=36182
    On EX40 this notification occurs on *any* kernel - we have tested this with 2.6.32, 3.0.88 and 3.10.4 kernels on stock Centos-6 64-bit. Perhaps the fix was not yet pushed into mainline kernel?

    Regardless, looks like the BIOS update did solve *something*, as we have had no crashes for the past 72 hours now (and we did not touch anything on the server and even went back to 2.6.32 kernel).
    Dan
    Check out our premium self-managed SSD VPS

Page 6 of 11 FirstFirst ... 3456789 ... LastLast

Similar Threads

  1. Hetzner billing issues
    By Munch1 in forum Web Hosting Lounge
    Replies: 5
    Last Post: 03-20-2012, 12:58 PM
  2. Hetzner experiences ?
    By SpaceStation in forum Dedicated Server
    Replies: 14
    Last Post: 01-09-2012, 05:33 AM
  3. Anyone had issues with hetzner later?
    By alex2012 in forum Dedicated Server
    Replies: 9
    Last Post: 12-15-2011, 02:06 PM
  4. Anyone having network issues with Hetzner?
    By ppwill in forum Dedicated Server
    Replies: 4
    Last Post: 07-02-2011, 04:46 PM
  5. Hetzner DC 12 Network Issues
    By Afterburst-Nick in forum Providers and Network Outages and Updates
    Replies: 1
    Last Post: 07-12-2010, 05:23 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •