Using compressed DASD files you can significantly reduce the file space required for emulated DASD files and possibly gain a performance boost because less physical I/O occurs. Both CKD (Count-Key-Data) and FBA (Fixed-Block-Architecture) emulation files can be compressed.
In regular (or uncompressed) files, each CKD track or FBA block occupies a specific spot in the emulation file. The offset of the track or block in the file can be directly calculated knowing the track or block number and the maximum size of the track or block. In compressed files, each track image or group of blocks may be compressed by zlib or bzip2, and only occupies the space neccessary for the compressed image. The offset of a compressed track or block is obtained by performing a two-table lookup. The lookup tables themselves reside in the emulation file.
Because FBA blocks are 512 bytes in length, and that being a rather small number, FBA blocks are grouped into block groups. Each block group contains 120 FBA blocks (60K).
Whenever a track or block group is written to a compressed file, it is written either to an existing free space within the file, or at the end of the file, then the lookup tables are updated, and then the space the track or block group previously occupied is freed. The location of a track or block group in the file can change many times.
In the event of a failure (for example, Hercules crash, operating system crash or power failure), the compressed emulation file on the host's physical disk may be out of sync if the host operating system defers physical writes to the file system containing the emulation file. Several methods have evolved to reduce the amount of data lost in these kind of events.
A compressed file may occupy only 20% of the disk space required by an uncompressed file. In other words, you may be able to have 5 times more emulated volumes using compressed DASD files. However, compressed files are more sensitive to failures and corruption may occur.
A compressed CKD or FBA dasd can have more than one physical file. The additional files are called shadow files. The function is implemented as a kind of snapshot, where a new shadow file can be created on demand. An emulated dasd is represented by a base file and 0 or more shadow files. All files are opened read-only except for the current file, which is opened read-write.
Shadow files are specified by the sf=shadow-file-name parameter on the device statement for the compressed DASD device.
Please note that the specified shadow filename does not have to
actually exist. The shadow-file-name operand of the sf=
parameter is simply a filename template that will be used
to name the shadow file whenever a shadow file is to be created, but
shadow files do not actually get created until you specifically create
them via the
sf+*) command. Please
refer to the discussion of the sf command several paragraphs
below for more information.
The shadow file name should have spot where the shadow file number will be set. This is either the character immediately preceding the filename extension or the last character of the file name itself if there is no extension.
For example, the following device statement:
0001 2311 VOL111.model-x.ext sf=VOL111_Shadow_0.model-x.extwill cause a shadow file with the following name to be created:
VOL111_Shadow_0.model-1.extwhereas the following slightly different device statement:
0002 2311 VOL222.model-x.ext sf=VOL222.model-x_Shadow_0.extwill cause a shadow file with the following name to be created:
Notice the placement of the shadow file number in the above two examples: it is always immediately before the last period of the filename. In earlier versions of Hercules the shadow file number was erroneously placed immediately before the first period of the filename instead of the last period. In newer versions of Hyperion this has been corrected to place the number immediately before the last period, as it was originally intended.
Using shadow files, you can keep the base file on a read-only device such as cdrom, or change the base file attributes to read-only, ensuring that this file can never be corrupted.
With the use of shadow files you can reconfigure your guest
without worrying about whether the new settings might render
your guest unusable. Simply create a new shadow file before
you do your reconfiguration. If something goes wrong, simply exit Hercules
and then delete that set of shadow files. You'll be right back to where
you were before you made your changes! Then you can simply try again.
Once you are satisfied that your changes are good, you can then do a
sf-*" backwards merge (see further below) to "commit" your changes.
There can be up to 8 shadow files in use at any time for an emulated dasd device. The base file is designated file and the shadow files are file to file. The highest numbered file in use at a given time is the current file, where all writes will occur. Track reads start with the current file and proceed down until a file is found that actually contains the track image.
Hercules console commands are provided to add a new shadow file, remove the current shadow file (with or without backward merge), compress the current shadow file, and display the shadow file status and statistics:
sfd unit Display shadow file status and statistics sfc unit Compress the current file sf+ unit Create a new shadow file sf- unit merge
Remove a shadow file.
If merge is specified or defaulted, then the contents of the current file is merged into the previous file, the current file is removed, and the previous file becomes the current file. The previous file must be able to be opened read-write. If nomerge is specified then the contents of the current file is discarded and the previous file becomes the current file. However, if the previous file is read-only, then a new shadow file is created (re-added) and that becomes the current file. The force option is required when doing a merge to the base file and the base file is read-only because the ro option was specified on the device config statement.
sfk unit level Perform the chkdsk function on the current file.
Level is a number from -1 ... 4, with the default level being 2:
-1 devhdr, cdevhdr, L1 table.
0 devhdr, cdevhdr, L1 table, L2 tables.
1 devhdr, cdevhdr, L1 table, L2 tables, free spaces.
2 devhdr, cdevhdr, L1 table, L2 tables, free spaces, trkhdrs. (default)
3 devhdr, cdevhdr, L1 table, L2 tables, free spaces, trkimgs.
4 devhdr, cdevhdr, build everything else from recovery.
You can use * in place of 'unit' in any of the above commands to apply the command to all compressed dasd. (
A compressed DASD file has 6 types of spaces, a device header, a compressed device header, a primary lookup table, secondary lookup tables, track or block group images, and free spaces. The first 3 types (device header, compressed device header, primary lookup table) only occur once, in order, at the very beginning of the file. The rest of the file is occupied by the other 3 space types.
It is also important to note that, except for the actual track or block group images space types, all numeric fields in each of the above space types are always kept in LITTLE endian format and are automatically converted to host format as needed before being used. This is the complete opposite of the way emulated guest storage is maintained (which is always in big endian format).
The very first 512 bytes of a compressed DASD file is the device header. The device header contains an "eye-catcher" that identifies the file type (32-bit or 64-bit CKD or FBA and base or shadow). The original CCKD design used 32-bit file offset values thus limiting file sizes to only 4GB or less.
The newer 64-bit design however, uses 64-bit offset values, thereby allowing compressed dasd image files to grow up to the theoretical maximum of 18,000,000 TB (18,000,000,000 GB). The actual maximum file size however, is limited by both the operating system and the format of whatever file system you are using. On Windows, using NTFS for example, the maximum file size is limited to only 16 TB (16,000 GB).
(*) Normal (uncompressed) FBA/FBA64 dasd image files do not have device headers. The first 512 bytes of a normal (uncompressed) FBA/FBA64 dasd image file is the actual first sector of the emulated FBA dasd device itself.
Device-id Dasd image type CKD_P370 Normal (uncompressed) CKD dasd image CKD_C370 Compressed CKD dasd image CKD_S370 Compressed CKD dasd Shadow file n/a* Normal (uncompressed) FBA** dasd image FBA_C370 Compressed FBA dasd image FBA_S370 Compressed FBA dasd Shadow file CKD_P064 Normal (uncompressed) CKD64 dasd image CKD_C064 Compressed CKD64 dasd image CKD_S064 Compressed CKD64 dasd Shadow file n/a* Normal (uncompressed) FBA64** dasd image FBA_C064 Compressed FBA64 dasd image FBA_S064 Compressed FBA64 dasd Shadow file
(**) Normal (uncompressed) FBA64 dasd image files are exactly identical to normal (uncompressed) FBA dasd image files. There is absolutely no distinction between the two types. Their formats are exactly identical.
The device type and file size information is specified in this header and, except for the eye-catcher (or device-id), this header is identical for both compressed CCKD/CCKD64 and compressed CFBA/CFBA64 images as well as for uncompressed CKD/CKD64 images. As mentioned just above however, it is not used for uncompressed FBA/FBA64 images.
|"CKD_C370" (device-id or "eye-catcher")||dh_heads||dh_trksize|
The dh_heads, dh_trksize and dh_highcyl values, being numeric, are always kept in little endian format.
The next 512 bytes contains the compressed device header.
The compressed device header contains file usage information, such as the amount of free space in the file.
Note that even though this control block is labeled "CCKD_DEVHDR" (or "CCKD64_DEVHDR" for 64-bit images), it actually applies to (is used for) compressed FBA images (CFBA/CFBA64) too, as well as for compressed CKD images (CCKD/CCKD64).
The num_L1tab, num_L2tab, cdh_cyls, cdh_size, cdh_used, free_off, free_total, free_largest, free_num, free_imbed, and cmp_parm values, being numeric, are always kept in little endian format.
After the compressed device header is the primary lookup table (also called the Level 1 table or L1tab.) Each 4 byte unsigned entry in the L1tab (8 byte unsigned entry for CCKD64) contains the file offset of a secondary lookup table (or level 2 table or L2tab).
The track or block group number being accessed divided by 256 gives the index into the L1tab. That is, each L1tab entry (CCKD_L1ENT or CCKD64_L1ENT) represents 256 tracks or block groups. The number of entries in the L1tab is dependent on the size of the emulated device.
Because each Level 1 table entry (CCKD_L1ENT or CCKD64_L1ENT) is a numeric value, they are of course always kept in little endian format.
Following the primary lookup table (L1tab), in no particular order, are: secondary lookup tables (L2tabs), track or block group images, and free spaces.
Each secondary lookup table (or L2tab), contains 256 eight byte entries (256 sixteen byte entries for CCKD64). The entry is indexed by the remainder of the track or block group number divided by 256.
Each L2 entry contains an unsigned 4 byte offset (unsigned 8 byte offset for CCKD64) to the track or block group image (L2_trkoff = file offset to where the track image or block group begins), an unsigned 2 byte track or block group length (the actual amount of data contained on the track or in the block group) and an unsigned 2 byte size of the track or block group image (i.e. how much space is actually available for use at that particular file offset).
The length is the amount of available space currently being consumed by the track or block group image, and the size is the amount of space actually available. The size of course may be greater than the length to prevent short free spaces from accumulating in the file.
Because each field in the Level 2 table (L2_trkoff, L2_len and L2_size) is a numeric value, they are of course always kept in little endian format.
A track image or block group image contains two fields: a 5-byte header and a variable amount of data that may or may not be compressed. The length field in the secondary lookup table entry (L2_len field shown just above) includes the length of both the 5-byte header as well as the track or block group data.
The 5-byte track or block group header contains a 1 byte compression indicator and 4 bytes that identify the track or block group. The format of the identifier depends on whether the emulated device is CKD or FBA:
|CKD Track Header|
For CKD, the 2 byte CC is the cylinder number of the track image and the HH is the head number. These numbers are stored in big-endian byte order. When the compression indicator byte is zero the 5 byte header is identical to the Home Address (or HA) for the track image.
The data --- which may or may not be compressed --- begins with the R0
count and ends with the end-of-track marker,
which is a count field containing
FFFFFFFFFFFFFFFF (eight hex x'FF's).
The Home Address plus the uncompressed track data comprise the track image.
|FBA Block Group Header|
For FBA, the 4 byte nnnnnnnn is the big-endian sector number of the block group (i.e. the sector or block number of the first block in the group). The block group data (the individual sectors in the block group) --- which may or may not be compressed (depending on the cmp flag) --- follows immediately after the block group header.
There is of course no end-of-track marker at the end of an FBA block group. (FBA devices do not have tracks.)
Please note that, while the FBA Block Group Header is the first part of the block group, it is not part of the compressed image. Only the sectors themselves (which immediately follow the Block Group Header) are actually compressed. FBA devices do not have Track Headers (Home Addresses). The FBA Block Group Header should be considered an internal structure and is never exposed to the guest.
The cmp compression indicator byte contains the value 0, 1 or 2. Any other value is invalid:
|0||Data is uncompressed|
|1||Data is compressed using zlib|
|2||Data is compressed using bzip2|
Free space contains a 4 byte offset to the next free space (8 byte offset for CCKD64), a 4 byte length of the free space (8 byte length for CCKD64), and zero or more bytes of residual data. The fb_offnxt and fb_len values, being numeric, are always kept in little endian format:
|0||1||2||3||4||5||6||7||8 . . .|
|fb_offnxt||fb_len||residual . . .|
The minimum length of a free space is 8 bytes. The free space chain is ordered by file offset and no two free spaces are adjacent. The chain is terminated when a free space has zero offset to the next free space. The compressed device header free_off field contains the offset to the first free space block.
The free space chain is read when the file is opened for read-write and written when the file is closed. The free space chain is maintained in storage while the file is opened.
A track or block group image is read while executing a channel program or by the readahead thread. An image has to be read before it is updated or written to. An image may be cached. If an image is cached, then the channel program may complete synchronously. This means that if all the data a channel program accesses is cached and Hercules does not have to perform physical I/O, then the channel program runs synchronously within the SSCH or SIO instruction in the CPU thread. All DASD channel programs are started synchronously. If a CCW in the channel program requires physical I/O then the channel program is interrupted and restarted at that CCW asynchronously in a device I/O thread.
All compressed devices share a common cache; the devices can be a mixture of FBA and/or CKD device types. Each cache entry contains a pointer to a 64K buffer containing an uncompressed track or block group image. If the track or block group image being read is not found in the cache, then the oldest (or least recently used or LRU) entry that is not busy is stolen. A cache entry is busy if it is being read, or last accessed by an active channel program, or updated but not yet written, or being written. If no cache entries are available then the read must enter a cache wait. When images are detected to be accessed sequentially then the readahead thread(s) may be signalled to read following sequential images.
When a cache entry is updated or written to, a bit is turned on indicating the cache entry has been updated. When a cache wait occurs, or (more likely) during space recovery, a cache flush is performed. When the cache is flushed, if any entries have the updated bit on, then the writer thread(s) are signalled. The writer thread selects the oldest cache entry with the updated bit on, compresses the image, and writes it to the file. The new image is written to a new space in the file and then the space previously occupied by the image is freed. In certain circumstances, the image may be written under stress. A stress write occurs when a reading thread is in a cache wait or when a high percentage of cache entries are pending write. In this circumstance, the compression parameters are relaxed to reduce the CPU requirements. An image written under stress is likely to take up more space than the same image written not under stress. The writer thread(s) run 1 nicer than the CPU thread(s); compression is a CPU intensive activity.
Space recovery is also called, somewhat inaccurately, garbage collection. The primary function of the space recovery thread, or garbage collector, is to keep the emulated compressed DASD files as small as possible. After all, that is the reason for using compressed DASD files in the first place.
When a track or block group image is written, it is written to a new location in the file. It is either written to an existing free space within the file or to the end of the file, increasing the size of the file. The space previously occupied by the image is freed, but it is not immediately available for space allocation requests. Instead, it is pending free space. It is assigned a pending value (typically 2) that is decremented each space recovery cycle (typically every 10 seconds). When the pending value reaches 0 then the space is available for allocation. This increases the chance that a track or block group image can be recovered in the event of a failure.
The space recovery routine relocates track or block group images towards the beginning of the file, causing free space to move towards the end of the file. When a free space reaches the end of the file, it 'falls off', that is, the file size is reduced.
Simply, the space recovery routine selects a space after a sufficiently large non-pending free space. It then reads and writes consecutive spaces using the normal cckd read and write routines. The space read will become pending free space and will hopefully be written to a non-pending free space occurring earlier in the file. Sometimes it is necessary to write the space later in the file to increase free space size earlier in the file. Left to itself, the space recovery routine will eventually remove all free space from the file. However, it is not intended to be a replacement for the cckdcomp utility; rather, the intent is to provide sufficient free space to prevent excessive file growth.
Another function performed by space recovery is to relocate L2 (secondary
lookup) tables towards the beginning of the file. This simplifies chkdsk
recovery and enables the chkdsk function to complete more quickly during initialization.
The cckd command and initialization statement can be used to affect cckd processing. The CCKD initialization statement is specified as a Hercules configuration file statement and supports the same options as the cckd command explained below.
|cckd||help||Display cckd help|
|cckd||stats||Display current cckd statistics|
|cckd||opts||Display current cckd options|
|cckd||opt=value||Set a cckd option. Multiple options may be specified,|
|separated by a comma with no intervening blanks:|
|comp=n||Compression to be used|
|compparm=n||Compression parameter to be used|
|debug=n||Turn CCW tracing debug messages on or off|
|freepend=n||Set the free pending value|
|fsync=n||Turn fsync on or off|
|gcint=n||Garbage collection interval|
|gcparm=n||Garbage collection parameter|
|gcstart=n||Start garbage collector|
|linuxnull=n||Check for null linux tracks|
|nosfd=n||Turn off stats report at close|
|nostress=n||Turn stress writes on or off|
|ra=n||Number of readahead threads|
|raq=n||Readahead queue size|
|rat=n||Number of tracks to readahead|
|trace=n||Number of trace table entries|
|wr=n||Number of writer threads|
Override the compression used for all cckd files. -1 (default) means
don't override the compression.
|compparm=n||Compression parameter. A value between -1 and 9. -1 means use the default
parameter. A higher value generally means more compression at the expense
of cpu and/or storage.
|debug=n||Enables or disables debug tracing. When enabled, additional CCKD trace
messages are displayed when CCW tracing is enabled for a CCKD dasd device.
The default is 0 (disable CCKD trace mesages).
You can specify 0 (disable debug tracing) or 1 (enable debug tracing).
|freepend=n||Specifies the free pending value for freed space. When a
track or block group image is written the space it previously occupied
is freed. This space will not be available for future
allocations until n garbage collection intervals have completed.
In the event of a catastrophic failure, previously written track or
block group images should be recoverable if the current image has
not yet been written to the physical disk. By default the value
is set to -1. This means that if fsync is specified
then the value is 1 otherwise it is 2. If 0 is specified then freed
space is immediately available for new allocations.
The default is -1.
You can specify a number between -1 and 4.
|fsync=n||Enables or disables fsync. When fsync is enabled, then
the disk emulation file is synchronized with the physical hard
disk at the end of a garbage collection interval (however, no more
often than 5 seconds). This means that if freepend is
non-zero then if a catastrophic error occurs then the emulated disks
should be recovered coherently. However, fsync may cause
performance degradation depending on the host operating system and/or
the host operating system level.
The default is 0 (fsync disabled).
You can specify 0 (disable fsync) or 1 (enable fsync).
|gcint=n||Number of seconds the garbage collector thread waits during an interval.
At the end of an interval, the garbage collector performs space recovery,
flushes the cache, and optionally fsyncs the emulation file.
(However, the file will not be fsynced unless at least 5
seconds have elapsed since the last fsync).
The default is 10 seconds.
You can specify a number between 0 and 60. A value of 0
disables the default automatic behavior of the garbage collector,
converting it to manual on-demand mode instead. When in manual mode,
the garbage collector will only run by specific demand via the 'gcstart' option,
and once started, will only run through one garbage collection cycle
and then exit, requiring you to manually start it again when desired.
In automatic mode (gcint > 0), the garbage collector will start itself
automatically and remain idle (but still running) between each garbage
collection cycle, automatically starting the next cycle every 'gcint'
|gcparm=n||A value affecting the amount of data moved during the garbage collector's
space recovery routine. The garbage collector determines an amount of
space to move based on the ratio of free space to used space in an
emulation file, and on the number of free spaces in the file. (The
garbage collector wants to reduce the free space to used space ratio
and the number of free spaces).
The value is logarithmic; a value of 8 means moving 28 the selected value, while a negative value similarly decreases the amount to be moved. Normally, 256K will be moved for a file in an interval. Specifying a value of +8 can increase the amount to 64M. At least 64K will always be moved. Interestingly, specifying a large value (such as +8) may not increase the garbage collection efficiency correspondingly.
The default is 0.
You can specify any number between -8 and +8.
|gcstart=n||If set to 1 then space recovery will become active on any emulated
disks that have free space. Normally space recovery will ignore emulated
disks until they have been updated.
The default is 0.
|linuxnull=n||If set to 1 then tracks written to 3390 cckd volumes that were
initialized with the -linux option will be checked if they
are null (that is, if all 12 4096 byte user records contain zeroes).
This is used by the dasdcopy utility.
The default is 0.
|nosfd=n||If set to 1 the shadow file status and statistics report
will not be displayed when the device is closed.
The default is 0, meaning the shadow file status and statistics
report will be displayed like normal when the device is closed.
|nostress=n||Indicates whether stress writes will occur or not. A track
or block group may be written under stress when a high percentage of
the cache is pending write or when a device I/O thread is waiting for
a cache entry. When a stressed write occurs, the compression algorithm
and/or compression parm may be relaxed, resulting in faster compression
but at the expense of a slightly larger compressed image.
If nostress is set to one, then a stressed situation is ignored. You would typically set this value to one when you want create the smallest emulation file possible in exchange for a possible performance degradation.
The default is 0.
You can specify 0 (enable stressed writes) or 1
(disable stressed writes).
|ra=n||Number of readahead threads. When sequential track or block group
access is detected, some number (rat=) of tracks or
block groups are queued (raq=) to be read by one of the
The default is 2.
You can specify a number between 1 and 9.
|raq=n||Size of the readahead queue. When sequential track or block group
access is detected, some number (rat= ) of tracks or
block groups are queued in the readahead queue.
The default is 4.
You can specify a number between 0 and 16 (a value
of zero disables readahead).
|rat=n||Number of tracks or block groups to read ahead when sequential access
has been detected.
The default is 2.
You can specify a number between 0 and 16 (a value
of zero disables readahead).
|trace=n||Number of cckd trace entries. You would normally specify a non-zero
value when debugging or capturing a problem in cckd code. When the
problem occurs, you should enter the k Hercules console command
which will print the trace table entries.
The default is 64.
You can specify a number between 0 and 200000.
Each entry represents 128 bytes. Normally, for debugging, I use
|wr=n||Number of writer threads. When the cache is flushed updated
cache entries are marked write pending and a writer thread is signalled.
The writer thread compresses the track or block group and writes the
compressed image to the emulation file. A writer thread is cpu-intensive
while compressing the track or block group and I/O-intensive while writing
the compressed image. The writer thread runs one nicer than
the CPU thread(s).
The default is 2.
You can specify a number between 1 and 9.
|These utilities are deprecated. Use the below dasdcopy / dasdcopy64 utility instead.|
|dasdcopy||[-options] ifile [sf=sfile] ofile|
|dasdcopy64||[-options] ifile [sf=sfile] ofile|
|Copy / convert dasd image files from one type to another.|
|Note: Any shadow file specified as input is merged into the output,|
|rendering the input shadow file obsolete once the copy completes.|
|cckdcdsk||[-v] [-f] [-ro] [-level] filename1 [filename2 ...]|
|cckdcdsk64||[-v] [-f] [-ro] [-level] filename1 [filename2 ...]|
|Check the integrity and repair / recover one or more damaged compressed files.|
|cckdcomp||[-v] [-f] [-level] filename1 [filename2 ...]|
|cckdcomp64||[-v] [-f] [-level] filename1 [filename2 ...]|
|Remove all free space from a compressed file or files.|
|cckdswap||[-v] [-f] [-level] filename1 [filename2 ...]|
|cckdswap64||[-v] [-f] [-level] filename1 [filename2 ...]|
|Change the endianess or byte-order of a compressed file or files|
Greg Smith email@example.com