| View previous topic :: View next topic |
| Author |
Message |
The_Great_Sephiroth
Veteran
Joined: 03 Oct 2014
Posts: 1602
Location: Fayetteville, NC, USA
|
 Posted: Mon Jan 05, 2015 4:03 am Post subject: |
 |
|
Well I have a similar setup (no portage stuff) on Debian systems and after years they no fragmentation due to keeping logs, caches, temp, etc to their own space. How is partitioning bad? I've never understood that argument. Heck, back int he days of DOS, you would purposely partition a bunch so you could use more of the drive due to using smaller cluster (block) sizes. Note that the partitions listed above have either a 1k or 2k block size, so they're more efficient at storage.
I have hit an issue, but I believe it is an ebuild being faulty. I tried to install X and it stops on media-libs/libjpeg-turbo, and throws an error during configuration.
| Code: |
configure: error: configuration problem: maybe object file format mismatch.
|
Every time. I cannot get past this. All partitions are under 5% usage at this point, both in terms of space and inodes, so I am not out of space. Everythign else built and installed fine. It ran for twelve hours, hit this package, and died. My log has nothing useful. What should I do?
_________________
Ever picture systemd as what runs "The Borg"? |
|
| Back to top |
|
 |
eccerr0r
Watchman
Joined: 01 Jul 2004
Posts: 9691
Location: almost Mile High in the USA
|
| Posted: Mon Jan 05, 2015 7:48 am Post subject: |
|
|
The amount of space wasted by a large block size (which is only 4KB on ext2/3/4) is nowhere near the inconvenience given by the example I had given earlier, and you could use something like reiserfs or something that has tail packing if absolutely necessary.
to help debug, the typical response: what is your emerge --info data which shows your CFLAGS, USE flags and GCC versions? As well as more lines of text just about the failure, including possibly the configure.log data.
_________________
Intel Core i7 2700K/Radeon R7 250/24GB DDR3/256GB SSD
What am I supposed watching? |
|
| Back to top |
|
|
The_Great_Sephiroth
Veteran
Joined: 03 Oct 2014
Posts: 1602
Location: Fayetteville, NC, USA
|
| Posted: Mon Jan 05, 2015 1:52 pm Post subject: |
|
|
Where would the configure.log be located? I can provide everything when I return home tonight.
_________________
Ever picture systemd as what runs "The Borg"? |
|
| Back to top |
|
|
Tony0945
Watchman
Joined: 25 Jul 2006
Posts: 5127
Location: Illinois, USA
|
| Posted: Tue Jan 06, 2015 2:33 pm Post subject: |
|
|
| Quote: | | the read head is forced to make multiple trips across the entire disk for trivial things |
A very good point, Ant. I maintain a separate /boot to protect it from accidental damage by a rogue program (or mis-typing), but /boot is only needed at boot time or when building the kernel. On some machines, I do and others I don't. I'm interested in separate swap. I fear that the swap file might be spread across the disk leading to inefficiency, OTOH, localizing in a partitition probably results in read head motion anyway. One single partition is attractive. Separate /home seems to be useful in a multi-user multi-computer environment such as when I worked at Motorola and each developer had a PC but our home directory and our tools were on a separate server.
Do you know of any studies regarding the speed advantages of a swap partition vs a swap file? |
|
| Back to top |
|
|
eccerr0r
Watchman
Joined: 01 Jul 2004
Posts: 9691
Location: almost Mile High in the USA
|
| Posted: Tue Jan 06, 2015 2:46 pm Post subject: |
|
|
A swap partition is faster indeed, no filesystem translation needed but for the most part, hitting swap is slow enough that any additional translation doesn't make a whole heck of a lot more difference - you won't notice.
If you thrash swap (versus letting it idle most of the time) to an SSD however, it *may* start making a difference, but in this case I'm not sure you want to kill your SSD this way...
_________________
Intel Core i7 2700K/Radeon R7 250/24GB DDR3/256GB SSD
What am I supposed watching? |
|
| Back to top |
|
|
depontius
Advocate
Joined: 05 May 2004
Posts: 3509
|
| Posted: Tue Jan 06, 2015 4:22 pm Post subject: |
|
|
As another data point, I just checked my server - a PIII with a 30G hard drive. As said, it's a server, so it's a pretty stripped installation. On the other hand, it's also my http-replicator server, so while /usr/portage/distfiles is routinely stripped to nothing, /var/cache/http-replicator is serving my entire home cluster.
| Code: | # fdisk -l /dev/sda
Disk /dev/sda: 28 GiB, 30000000000 bytes, 58593750 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: dos
Disk identifier: 0x0001d1b0
Device Boot Start End Blocks Id System
/dev/sda1 2048 526335 262144 83 Linux
/dev/sda2 526336 58593749 29033707 5 Extended
/dev/sda5 528384 2625535 1048576 82 Linux swap / Solaris
/dev/sda6 2627584 58593749 27983083 83 Linux |
Root is getting pretty (86%) full, but /var/cache/http-replicator is quite prone to accumulating cruft. I need to clean it out.
I'm also serving /home over nfsv4, but that's on a separate RAID-1 pair. My IMAP maildirs are also on the RAID-1 pair.
_________________
.sigs waste space and bandwidth |
|
| Back to top |
|
|
The_Great_Sephiroth
Veteran
Joined: 03 Oct 2014
Posts: 1602
Location: Fayetteville, NC, USA
|
| Posted: Tue Jan 06, 2015 7:42 pm Post subject: |
|
|
I figured out the issue. Since I copied the configuration from my system and changed the kernel options and such I thought I was good. Incorrect! I left CHOST set to x86_64-*. I changed it to i686-*, but am going to reinstall to be safe. I may change some things in regards to partitioning also.
_________________
Ever picture systemd as what runs "The Borg"? |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
Installing Gentoo
